Architectural and Civil Drafters
17-3011.00Median wage $66,150/yr103,700 employed (US)Rank #135 of 923 scored · top 15% by substitution
Prepare detailed drawings of architectural and structural features of buildings or drawings and topographical relief maps used in civil engineering projects, such as highways, bridges, and public works. Use knowledge of building materials, engineering practices, and mathematics to complete drawings.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution42
Exposure41
Augmentation67
Substitution — the headline: capability discounted by cost, barriers and adoption.
Exposure — technical capability alone, regardless of whether anyone deploys it.
Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.
Tasks on the substitution scale
28 rated tasks, binned by substitution score.
Position among all scored occupations
Distribution of 923 occupation scores; the marker is this occupation.
Tasks with substitution ≥ 70
11%
Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.
Why this score
The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.
Task automatabilityw 35%44
panel mean rating 2.8/5 → substitution pressure 44/100
Technical feasibility todayw 20%37
panel mean rating 2.5/5 → substitution pressure 37/100
Cost vs. human wagew 15%39
panel mean rating 2.6/5 → substitution pressure 39/100
Adoption barriersw 20%inverted — strong barriers lower the score51
panel mean rating 3.0/5 (barrier strength) → substitution pressure 51/100
Sector adoption velocityw 10%38
panel mean rating 2.5/5 → substitution pressure 38/100
Task breakdown (28 tasks)
Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.
Reproduce drawings on copy machines or trace copies of plans and drawings, using transparent paper or cloth, ink, pencil, and standard drafting instruments.
100CI 100–100 · exposure 100 · augmentation 38 · importance 2.6/5 · click for rater detail
Reproduce drawings on copy machines or trace copies of plans and drawings, using transparent paper or cloth, ink, pencil, and standard drafting instruments.
100| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Digital document workflows and automated copiers have been standard practice in architectural and engineering firms for two decades; the shift from manual tracing to scanning and digital file sharing is near-universal in the sector. |
| Sector adoption velocity | claude-sonnet-5 | 5/5 | Architectural and engineering drafting fully digitized decades ago with CAD/BIM tools and digital file management, making manual tracing essentially obsolete in practice. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Once a drafter has a final design, machines handle reproduction with no human assistance needed; augmentation potential is minimal because the task itself—pure copying—offers little room for AI-human collaboration beyond the machine completing it autonomously. |
| Augmentation potential | claude-sonnet-5 | 3/5 | While the original manual task itself is largely replaced rather than augmented, digital tools do assist drafters in managing, duplicating, and revising drawings efficiently. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Reproducing drawings via photocopying and tracing with pencil/ink on transparent media is a straightforward mechanical task that modern document scanning, digital image processing, and printing systems handle end-to-end with negligible time overhead and superior quality compared to manual tracing. |
| Task automatability | claude-sonnet-5 | 5/5 | This is a purely mechanical reproduction/tracing task with no design judgment involved; digital scanning, CAD copy/paste, and file duplication fully replace manual reproduction with far more than 50% time savings at equal or better quality.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 1/5 | No licensing, regulatory, or human-contact requirement exists for copying or scanning drawings; organizations face no legal or contractual barriers to substituting machine copying for manual reproduction. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, liability, or regulatory requirement mandates manual reproduction; it's purely a legacy manual process now handled digitally. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated photocopying and digital scanning cost pennies per page; full-stack cost (hardware amortization, consumables, minimal oversight) is orders of magnitude lower than paying a drafter's loaded wage (typically $50–80/hour) to manually trace or copy drawings. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Digital duplication and scanning cost pennies compared to the labor time of manually tracing or operating copy machines for the same output. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Multifunction copiers and document imaging systems are ubiquitous in production use; automated scanning-to-print workflows and digital file conversion are mature, deployed at scale across organizations worldwide, and reliably perform plan reproduction and duplication. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | Standard scanners, plotters, and CAD software have replaced manual tracing and copy-machine reproduction in virtually all production drafting environments for decades. |
Finish and duplicate drawings and documentation packages according to required mediums and specifications for reproduction, using blueprinting, photography, or other duplicating methods.
92CI 87–97 · exposure 95 · augmentation 75 · importance 4.0/5 · click for rater detail
Finish and duplicate drawings and documentation packages according to required mediums and specifications for reproduction, using blueprinting, photography, or other duplicating methods.
92| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Architecture and engineering firms have rapidly adopted CAD software, document management systems, and automated batch processing over the past decade. Digital-first workflows are now standard in most professional design practices. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | AEC firms have broadly adopted digital CAD workflows and automated plotting/export tools, though full end-to-end adoption varies by firm size and legacy systems. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools assist drafters by automating repetitive finishing tasks, standardizing document formatting, and managing reproduction workflows, allowing humans to focus on design review and quality control rather than manual duplication labor. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Software greatly speeds up formatting, batch exporting, and reproduction tasks, letting drafters focus on higher-value design work while automation handles routine finishing. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | AI can fully automate finishing, duplicating, and converting architectural drawings across digital mediums (PDF, CAD formats, image formats) with >50% time savings. Current tools handle document standardization, format conversion, and batch processing of technical drawings end-to-end. |
| Task automatability | claude-sonnet-5 | 5/5 | Finishing and duplicating drawings for reproduction (formatting, plotting, exporting to required file types/mediums) is a highly standardized, repetitive digital task that CAD/PDM software and scripts already automate almost entirely. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers exist; drawings are assets owned by firms, and duplication is a technical process with no licensing requirement or mandatory human sign-off. Organizational adoption is the main friction, not legal restriction. |
| Adoption barriers | claude-sonnet-5 | 1/5 | This is a mechanical/administrative finishing step with no licensing, liability, or human-contact requirement attached to the reproduction process itself. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated digital duplication and format conversion costs (server processing, storage, software licensing) are orders of magnitude cheaper than manual finishing and physical reproduction labor, especially for large batches. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated plotting/export/duplication pipelines cost a tiny fraction of a drafter's hourly wage once set up, and are already standard in most firms. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products exist (CAD software with batch processing, document management systems, format converters) that reliably perform drawing duplication and format conversion in production. Minor gaps remain in complex edge cases or legacy blueprint specifications, but core task is deployable at scale. |
| Technical feasibility today | claude-sonnet-5 | 5/5 | CAD systems, plotting software, and document management systems reliably batch-process, format, and reproduce drawings in production environments across the AEC industry today. |
Calculate heat loss and gain of buildings and structures to determine required equipment specifications, following standard procedures.
75CI 72–77 · exposure 75 · augmentation 100 · importance 2.3/5 · click for rater detail
Calculate heat loss and gain of buildings and structures to determine required equipment specifications, following standard procedures.
75| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Architectural and engineering firms have rapidly adopted building energy modeling and integrated CAD-linked load calculation tools over the past 10–15 years, driven by energy code compliance and competition. This is one of the most automated task domains in construction-adjacent professions. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Architecture/engineering firms have moderate digitization with common use of calculation software and BIM tools, but full end-to-end automation adoption is still uneven across firms of varying size. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI-powered load calculators and parametric energy models dramatically augment drafter productivity by automating iteration, sensitivity analysis, and specification generation. Drafters use these tools to explore design variants and validate assumptions far faster than manual calculation, keeping humans fully in the loop while multiplying output quality. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI-enabled calculation and energy modeling tools substantially speed up and improve accuracy of heat loss/gain determinations, letting drafters focus on design integration and verification. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | This task involves standard thermodynamic calculations, load analysis, and equipment sizing that can be largely automated using building energy simulation software and rule-based tools. Modern CAD/BIM systems and HVAC load calculation software (e.g., Carrier HAP, Trane TRACE) automate 70–80% of the computational work, though human validation of inputs and interpretation of edge cases remain necessary for time-to-completion equivalence. |
| Task automatability | claude-sonnet-5 | 4/5 | Heat loss/gain calculations follow standardized formulas (Manual J, ASHRAE methods) that are highly codifiable and already handled by specialized software; AI/software tools can perform these calculations with minimal human input given building specs.dorm |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Professional liability and code compliance create modest friction: engineers must typically sign off on calculations and equipment specs, so complete substitution is limited. However, automation of intermediate calculation steps faces no legal or licensing barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement mandates a human perform the raw calculation itself, though final specifications often require review/sign-off by a licensed engineer for liability reasons. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI/software tools cost a fraction of a full-time drafter's loaded salary per calculation instance. A single thermal analysis run costs $50–200 in software and compute time versus $150–300 in billable drafter labor, delivering an approximate 2–5× cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated calculation software is inexpensive per calculation compared to a drafter's hourly wage, though initial data entry/model setup still requires some labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature, production-grade software tools exist and are widely deployed in architectural/engineering firms for thermal load calculation. Products like Revit with energy analysis plugins, OpenStudio, and dedicated HVAC tools reliably perform standard heat loss/gain calculations at scale, though integration with downstream specification requires human review. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature engineering software (e.g., Manual J calculators, energy modeling tools like eQuest, IES-VE) already automate this reliably in production, though some require input curation and professional review. |
Check dimensions of materials to be used and assign numbers to lists of materials.
70CI 67–72 · exposure 70 · augmentation 75 · importance 3.9/5 · click for rater detail
Check dimensions of materials to be used and assign numbers to lists of materials.
70| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Architecture and construction are moderately digitized and adopt automation unevenly. While some design firms and large contractors have adopted AI-assisted material tracking, widespread production adoption is still emerging rather than mature. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | AEC (architecture/engineering/construction) firms are moderate adopters of BIM/CAD automation tools with growing use of parametric and rule-based checking, but overall sector digitization lags fast-moving software/finance sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can assist drafters significantly by auto-populating dimensions, flagging spec mismatches, and suggesting material assignments, substantially accelerating the checklist-and-assignment workflow while the human retains oversight of material selections. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted CAD tools significantly speed up dimension verification and materials list generation, letting drafters focus on design judgment while software handles tedious counting and numbering. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI vision systems can reliably identify and measure materials from images or technical specifications, and automated systems can cross-reference dimensions against standardized lists with minimal error. This task is largely procedural—extract dimensions, verify against specs, and assign catalog numbers—which is well within current LLM and computer vision capabilities, easily achieving >50% time savings. |
| Task automatability | claude-sonnet-5 | 4/5 | This is a structured, rules-based verification and cataloging task involving comparing dimensions and assigning identifiers, which is well within reach of CAD-integrated software and AI tools that can cross-check dimensions and auto-generate material lists/schedules. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers exist; the task is primarily clerical and verification-based, not design-critical sign-off. Some organizations may require human spot-checks or integration with existing CAD workflows, but nothing legally mandates human performance. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically for this sub-task; some organizational friction exists since final drawings may need drafter/engineer sign-off but the numbering task itself has few hard barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference cost for dimension verification and assignment is typically $0.01–$0.10 per item, far cheaper than paying a drafter ($25–$50/hour loaded) for the same work. Overhead is minimal—no specialized training or equipment required. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated BOM generation and dimension-checking scripts within CAD platforms run at marginal compute cost versus paying a drafter's hourly wage for manual list assignment and checking. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed computer vision and data-processing tools (including GPT-4V and similar systems) can perform dimension checking and material assignment in production settings. While some edge cases (ambiguous materials, non-standard specs) may require human review, mature products reliably handle routine dimension verification and numbering workflows at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | CAD software (Revit, AutoCAD) already auto-generates material schedules and bill-of-materials with dimension checks, but fully autonomous dimension verification against real-world tolerances and error-free numbering still requires human review in production settings. |
Calculate weights, volumes, and stress factors and their implications for technical aspects of designs.
63CI 45–81 · exposure 62 · augmentation 100 · importance 3.1/5 · click for rater detail
Calculate weights, volumes, and stress factors and their implications for technical aspects of designs.
63| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Architectural and engineering sectors have been deeply digitized for decades, with CAD and FEA software standard practice in most firms, indicating fast and sustained adoption of automated calculation tools. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | AEC (architecture/engineering/construction) industries show moderate digitization with growing BIM and computational design tool adoption, but remain slower than pure information/finance sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI-powered parametric design and real-time stress visualization dramatically assist drafters by instantly showing design implications, enabling rapid iteration and exploration that would be infeasible with manual calculation, keeping the human in active decision-making. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI and parametric design tools significantly speed up calculation and iteration for drafters, letting them explore more design options while retaining responsibility for final engineering judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI and CAD software can automatically calculate weights, volumes, and basic stress factors from 3D models using finite element analysis and parametric design tools, achieving substantial time savings. However, interpreting implications for complex design trade-offs still requires human judgment, preventing a full 5-rating. |
| Task automatability | claude-sonnet-5 | 3/5 | Standard weight/volume/stress calculations can be automated via CAD-integrated engineering software and AI-assisted tools, but interpreting results and applying engineering judgment for novel or complex designs still requires human oversight, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard regulatory or licensing barriers exist for the calculation itself; firms adopt these tools routinely. However, some jurisdictions require licensed engineers to stamp or review structural calculations, creating modest friction but not preventing automation of the computation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Structural calculations often require sign-off by a licensed engineer or architect for safety and code compliance, creating strong liability-driven barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Software licenses and computational inference are orders of magnitude cheaper than hiring a drafter or structural analyst for each calculation cycle, especially when amortized across many projects and reiterations. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Software licenses and computation are cheap, but integration, calibration to codes/standards, and required professional review add substantial overhead comparable to a drafter's efficient use of existing tools. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature CAD and structural analysis software (AutoCAD, CATIA, Ansys, Revit) reliably perform these calculations in production environments across architecture and engineering firms. These tools are industry-standard and demonstrate consistent performance at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Structural analysis software (e.g., Revit, AutoCAD with add-ins, finite element tools) reliably performs many calculations in production, but these are specialized engineering tools rather than general AI systems, and complex stress analysis still needs expert validation. |
Draw maps, diagrams, and profiles, using cross-sections and surveys, to represent elevations, topographical contours, subsurface formations, and structures.
62CI 46–79 · exposure 62 · augmentation 88 · importance 4.2/5 · click for rater detail
Draw maps, diagrams, and profiles, using cross-sections and surveys, to represent elevations, topographical contours, subsurface formations, and structures.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Engineering, surveying, and construction firms have rapidly adopted AI-assisted CAD and GIS workflows over the past 3–5 years; major software vendors now ship automated drafting as standard features, and displacement is measurable in production pipelines. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Civil/architectural firms are adopting BIM and automated drafting tools steadily, but the sector overall (construction-adjacent) lags behind fast-moving digital-native industries. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically accelerates the drafting workflow by automating routine geometry and rendering tasks, allowing human drafters to focus on interpretation, design refinement, and regulatory compliance—a textbook example of high-value human-AI partnership. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI and automated CAD tools significantly speed up generation of contours, cross-sections, and profiles from survey data, greatly aiding drafters while they verify and finalize outputs. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can automate most of the technical drawing process: converting survey data into maps, generating cross-sections, and rendering topographical contours with minimal human intervention. However, some interpretive decisions about feature representation and validation against site-specific requirements typically require human review, preventing a full 5. |
| Task automatability | claude-sonnet-5 | 3/5 | AI-assisted CAD/GIS tools can generate contour lines, cross-sections, and profiles from survey data, but integrating varied field data, ensuring accuracy, and finalizing drawings still require significant drafter setup and judgment.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Professional liability and regulatory sign-off typically require a licensed engineer or surveyor to verify and stamp drawings, but the automation itself faces no legal prohibition. Most barriers are organizational (quality control, client expectations) rather than statutory. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement specifically for drafting itself, but engineering sign-off and liability for structural/civil accuracy create moderate friction against pure automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven CAD and GIS tools cost a fraction of a drafter's loaded wage per drawing once data is ingested; inference and integration are highly commoditized in the geospatial software stack. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Software licenses plus need for skilled oversight to validate survey-derived drawings keep costs from falling dramatically below human drafting costs, though some efficiency gains exist. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature CAD and GIS software with AI plugins (AutoCAD, ArcGIS) now routinely perform automated map and diagram generation from survey data in production environments. Products reliably generate profiles and contours, though quality assurance and customization remain partially manual. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | CAD software with automation features (Civil 3D, AutoCAD add-ins) and AI-assisted plugins exist and are used in production, but full end-to-end automation of complex topographic/subsurface drawing still has notable error rates and requires human correction. |
Prepare colored drawings of landscape and interior designs for presentation to client.
61CI 47–75 · exposure 58 · augmentation 88 · importance 3.2/5 · click for rater detail
Prepare colored drawings of landscape and interior designs for presentation to client.
61| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Architecture, interior design, and landscape design firms are actively adopting generative visualization tools in production workflows. Many studios now use AI for client presentations and concept iteration, representing fast mainstream adoption in a digitized professional sector. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Architecture and design firms are experimenting with AI visualization tools but production integration remains limited compared to faster-moving sectors like finance or general office work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically augments drafter productivity by generating multiple design options, rapid iterations, and client-ready renderings that human designers then refine and customize. This assistive use is already widespread, with designers spending less time on manual rendering and more on creative direction. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI rendering and coloring tools meaningfully speed up the visualization process, letting drafters iterate faster on presentation options while retaining control over technical accuracy and final design intent. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can generate colored drawings of landscape and interior designs with high quality and substantial time savings. Generative models can produce presentation-ready renderings from text descriptions or sketches, reducing manual drafting time by well over 50%, though human direction and refinement typically remains. |
| Task automatability | claude-sonnet-5 | 3/5 | AI image generation tools can produce colored renderings from descriptions or rough plans, but achieving precise, technically accurate interior/landscape designs matching client specifications still requires significant human drafting and revision.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory barriers exist for AI-generated presentation drawings. However, professional liability concerns, client preference for human-crafted work, and internal QA/review workflows create moderate friction to full substitution. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically for colored presentation drawings, though client trust and design accountability create some preference for human-produced work tied to a project's technical documentation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-generated renderings cost orders of magnitude less than hiring a drafter to produce colored presentation drawings manually. A single inference run and refinement cycle costs dollars, versus hours of human labor at professional rates. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI rendering can cut time on visualization drafts, but human oversight, revisions, and integration with technical drawings keep costs roughly comparable to a drafter's output once quality and accuracy are ensured. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (Midjourney, DALL-E, Stable Diffusion, and specialized architecture tools like Spacemaker, Architailor) reliably produce colored interior and landscape design visualizations in production. Performance is strong for presentation purposes, though some organizations still prefer or require human-led iteration. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI rendering tools (e.g., Midjourney, architectural visualization AI) exist and are used for concept art, but production-grade CAD-integrated colored presentation drawings for client-ready deliverables still largely rely on drafters using specialized software with AI as a supplement. |
Determine quality, cost, strength, and quantity of required materials, and enter figures on materials lists.
59CI 50–67 · exposure 58 · augmentation 75 · importance 3.5/5 · click for rater detail
Determine quality, cost, strength, and quantity of required materials, and enter figures on materials lists.
59| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | BIM and CAD software adoption is widespread in architecture and construction, and material takeoff automation is increasingly common in larger firms and projects, but smaller practices and regional variations show slower uptake. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | AEC/construction sector has moderate BIM and takeoff-tool adoption with growing but still uneven use of AI-driven quantity extraction across firms. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly accelerates the data entry and lookup phases of materials list creation, allowing drafters to focus on verification, site-specific adjustments, and complex material selections; productivity gains are substantial while human judgment remains valuable. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted takeoff and cost-estimation tools meaningfully speed up drafters' quantity and cost calculations while drafters retain responsibility for final specification decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can reliably extract and calculate material specifications from architectural drawings, cross-reference building codes and standards, and populate materials lists with high accuracy. Human oversight for final review is typically still required, but the computational and lookup portions achieve well over 50% time savings. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can extract quantities and generate material lists from BIM/CAD models and cross-reference cost databases, but judgment on quality/strength specification still needs engineering validation, so only part of the task is fully automatable today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No legal or licensing requirement mandates that a human must compile materials lists; the main friction is organizational inertia and the need for quality assurance oversight, which only partially constrains adoption. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for material listing itself, but liability for incorrect strength/quantity specs on load-bearing elements creates professional oversight friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-based material takeoff tools cost a fraction of the labor for manual quantity surveying and material list compilation, particularly when integrated into existing BIM workflows; all-in costs are substantially lower than hiring a drafter for this task. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Automated takeoff tools reduce labor time substantially but still require licensed software, integration, and human review, making costs meaningfully lower but not an order of magnitude cheaper overall. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | CAD-integrated AI tools and BIM software with material quantification features exist in production (e.g., Revit plugins, specialized quantity takeoff software), but accuracy and coverage remain variable across complex projects; many firms still rely on manual checking or hybrid workflows. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | BIM software (Revit, quantity takeoff tools) with AI-assisted extraction is deployed in production, but accuracy issues and need for manual verification of specs and cost data are common. |
Produce drawings, using computer-assisted drafting systems (CAD) or drafting machines, or by hand, using compasses, dividers, protractors, triangles, and other drafting devices.
56CI 43–70 · exposure 58 · augmentation 88 · importance 4.9/5 · click for rater detail
Produce drawings, using computer-assisted drafting systems (CAD) or drafting machines, or by hand, using compasses, dividers, protractors, triangles, and other drafting devices.
56| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Architecture and engineering firms are actively adopting CAD AI tools and generative design features, with fast diffusion in larger practices and professional service firms. Production deployment of AI-assisted drafting is increasingly common, though smaller practices lag. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Architecture/engineering/construction sectors are historically slower digitization adopters compared to pure information sectors, with AI drafting tools still in early pilot stages. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically augments drafter productivity by automating repetitive linework, generating views and details from models, and accelerating revision cycles. Drafters remain in the loop for design validation and quality control, making this a high-productivity augmentation scenario. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI significantly speeds up repetitive drafting tasks, symbol placement, dimensioning, and design iteration, letting drafters focus on judgment-heavy aspects while maintaining oversight. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | CAD-based drawing production can be substantially automated today. AI can generate technical drawings from specifications, convert sketches to CAD, and produce standard views and details with high fidelity. While complex spatial reasoning and design intent may still require human input, the core drawing production task (linework, dimensioning, layering) meets the ≥50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 3/5 | AI tools can generate draft CAD layouts, floor plans, and basic technical drawings from specifications, but complex, code-compliant architectural/civil drawings still require substantial human refinement and verification. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Some organizational and professional friction exists: drafters are embedded in design teams, drawings often require human judgment about intent and compliance, and liability concerns may require human sign-off. However, there is no legal requirement that a licensed professional must personally produce the drawing lines, creating moderate (not hard) friction. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While drafting itself isn't strictly licensed, drawings often require sign-off by licensed architects/engineers, and liability for errors in construction documents creates meaningful oversight requirements. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-based drawing generation has very low marginal inference cost compared to the loaded wage of a drafter (typically $50–70k annually). Integration and oversight add modest cost, but the all-in ratio favors AI substantially, especially for repetitive or template-based drawing tasks. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted drafting tools can reduce time on repetitive elements, but integration, licensing, and necessary human review keep costs roughly comparable to skilled drafter labor for complete deliverables. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (Autodesk's generative design features, AI-assisted drafting plugins, sketch-to-CAD converters) reliably perform portions of this task in production environments. The technology is mature for standard drawing types, though custom or highly specialized drawings may still require human oversight and refinement. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some CAD plugins and generative design tools exist (e.g., AI-assisted layout generators), but production use for full, accurate technical drawings remains limited and requires heavy human correction. |
Calculate excavation tonnage and prepare graphs and fill-hauling diagrams for use in earth-moving operations.
54CI 39–70 · exposure 58 · augmentation 88 · importance 3.3/5 · click for rater detail
Calculate excavation tonnage and prepare graphs and fill-hauling diagrams for use in earth-moving operations.
54| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Architecture and civil engineering firms are digitally mature sectors with high adoption of CAD and BIM tools; incremental adoption of AI-assisted calculation and diagramming is already occurring in practice, particularly among larger firms. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Construction and civil engineering sectors are historically slow adopters of AI-driven automation compared to purely digital/informational industries, with adoption concentrated in large firms using specialized software rather than broad AI agents. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically augments drafter productivity by automating tedious calculations and generating initial diagrams in seconds, leaving the human to focus on design validation, quality review, and site-specific adjustments. This is already a high-impact augmentation workflow in professional practice. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enhanced CAD tools and calculation software significantly speed up tonnage estimates and diagram generation, letting drafters focus on verification and refinement rather than manual computation. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Excavation tonnage calculations are largely mathematical and geometric; current AI can extract site data, apply standard soil density formulas, and generate graphs/diagrams with high accuracy. The task requires structured inputs (volume, soil type, density) which are increasingly digitized, and produces standardized outputs (graphs, diagrams) that AI systems excel at generating today. |
| Task automatability | claude-sonnet-5 | 3/5 | Excavation tonnage calculation and haul diagrams involve structured, formula-based computation that AI/CAD-integrated tools can partially automate, but require accurate site data input and civil engineering judgment that limits full end-to-end automation today.atinum. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Professional practice standards require a licensed engineer or drafter to review and certify diagrams for construction use, creating a human sign-off requirement. Quality oversight and integration into established project workflows introduce moderate friction, though the calculations themselves face no hard legal barriers. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not always requiring a licensed professional to draft, civil engineering calculations often need review/sign-off by a licensed engineer for regulatory compliance, creating moderate liability and oversight barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference and integration costs for calculation and visualization are minimal compared to paying a drafter several hours per project for manual computation and drafting. Once set up, the per-task cost is orders of magnitude lower than human labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized civil engineering software licenses plus the need for skilled oversight to validate volumetric and haul calculations keep costs from being dramatically lower than a drafter's time, though some efficiency gains exist. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature CAD and engineering software (AutoCAD, Civil 3D) already automate volume calculations and diagram generation. AI-augmented tools and agents can parse site survey data, compute tonnage, and produce fill-hauling diagrams reliably in production environments used by engineering firms. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some earthworks/civil software (e.g., AutoCAD Civil 3D, Trimble) offers automated cut-fill and haul calculations, but these are specialized engineering tools rather than general AI products, and reliable deployment requires accurate survey data and human verification. |
Draw rough and detailed scale plans for foundations, buildings, and structures, based on preliminary concepts, sketches, engineering calculations, specification sheets, and other data.
47CI 39–55 · exposure 53 · augmentation 75 · importance 4.0/5 · click for rater detail
Draw rough and detailed scale plans for foundations, buildings, and structures, based on preliminary concepts, sketches, engineering calculations, specification sheets, and other data.
47| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Major architecture and engineering firms are piloting AI-assisted CAD and generative design, but production adoption remains inconsistent; smaller firms lag significantly, and cultural preference for human-validated designs slows penetration. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture/engineering/construction) firms have been slower than software/finance sectors to adopt AI in core production drafting workflows, though BIM automation is growing. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools substantially augment drafter productivity by rapidly generating initial geometry, automating routine annotation, and iterating design variants, while the drafter remains central to refinement, compliance checking, and client communication. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted CAD tools significantly speed up rough drafting, layer management, and repetitive detailing tasks, letting drafters focus on judgment-heavy design decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can now generate detailed architectural drawings from sketches and specifications using tools like generative design software and CAD AI assistants, achieving significant time savings (>50%) on routine plan generation, though human review and modification of complex or site-specific details remains necessary. |
| Task automatability | claude-sonnet-5 | 3/5 | AI-assisted CAD tools can generate rough layouts and detailed drawings from parametric inputs and sketches, but translating engineering calculations and specifications into compliant, precise scale plans still requires significant human review and iteration., especially for complex structures. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Professional licensing (PE/architect sign-off) and liability for design errors create moderate friction; clients often prefer human drafters for complex projects; regulatory code compliance still requires expert human judgment, limiting full substitution. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While drafters typically aren't licensed like engineers, output must align with building codes and engineering sign-off requirements, creating moderate liability and quality-control friction against pure automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI systems (software licensing, compute, integration, and mandatory human review) remain roughly comparable to or sometimes exceed the cost of an experienced drafter, especially when accounting for rework and verification labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Software licensing, integration with BIM/CAD systems, and required human verification of technical accuracy keep AI-assisted drafting costs comparable to or only modestly cheaper than skilled drafter labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products like Autodesk's generative design tools and AI-assisted CAD systems exist and are used in production, but they typically require substantial human input, iteration, and oversight; they handle standard geometries well but struggle with nuanced site constraints and code compliance verification. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some CAD plugins and generative design tools exist (e.g., AI-assisted BIM features), but they are narrow in scope and not yet reliably producing full construction-ready drawings without heavy drafter oversight. |
Lay out and plan interior room arrangements for commercial buildings, using computer-assisted drafting (CAD) equipment and software.
41CI 34–49 · exposure 45 · augmentation 75 · importance 4.1/5 · click for rater detail
Lay out and plan interior room arrangements for commercial buildings, using computer-assisted drafting (CAD) equipment and software.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architecture and construction sectors adopt digital tools slowly relative to software/finance. Most firms use traditional CAD workflows; generative layout tools remain niche pilots rather than standard production practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture/engineering/construction) sectors are historically slow to adopt AI tools relative to software/finance, with pilots more common than production deployment for layout generation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist drafters by generating layout options, checking code compliance, and automating repetitive constraint modeling, allowing the human to focus on design intent and client communication. This augmentation is already demonstrated in modern CAD ecosystems. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted CAD features, generative layout suggestions, and automation of repetitive drafting tasks meaningfully speed up drafters' workflow while they retain control over design decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can generate initial layout proposals and automate constraint-checking (egress, accessibility, code compliance), but final spatial decisions require understanding client intent, aesthetic judgment, and approval workflows. Roughly half the iterative design work could be automated with significant CAD integration setup. |
| Task automatability | claude-sonnet-5 | 3/5 | AI tools can generate draft space-planning layouts from requirements and constraints, but final interior arrangements for commercial buildings require iterative client input, code compliance checks, and spatial judgment that still need substantial human refinement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Building code compliance, liability for safety violations, and need for licensed architect or engineer sign-off on commercial buildings create substantial legal barriers. Client relationships and approval workflows also require human judgment and accountability. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement mandates a human specifically for drafting, but liability for code compliance and coordination with architects/engineers creates moderate organizational and professional friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI layout tools require expensive CAD licenses, integration engineering, and senior drafter oversight to validate outputs. The human cost savings are modest because human review and iteration remain the bottleneck. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Current AI-assisted layout tools still require licensed software, integration, and significant drafter oversight, so cost savings are modest rather than order-of-magnitude given verification and code-compliance needs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | CAD-integrated layout tools exist (e.g., space-planning plugins, generative design in Revit, AI-assisted sketching), but they struggle with novel constraints, client preferences, and code edge cases. Products work in controlled scenarios but require substantial human oversight and refinement in production. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some generative design and AI-assisted CAD plugins exist (e.g., space planning add-ons), but they are narrow, require heavy human correction, and are not yet standard production tools replacing drafters' layout work. |
Draft plans and detailed drawings for structures, installations, and construction projects, such as highways, sewage disposal systems, and dikes, working from sketches or notes.
37CI 34–41 · exposure 45 · augmentation 75 · importance 4.6/5 · click for rater detail
Draft plans and detailed drawings for structures, installations, and construction projects, such as highways, sewage disposal systems, and dikes, working from sketches or notes.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Large architecture and engineering firms have adopted CAD and parametric tools widely, but AI-driven automation of the full drafting workflow is still in pilot and early adoption phases. Smaller firms and specialized sectors lag significantly; adoption is uneven rather than rapid and deep. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture/engineering/construction) sector has historically lagged in digitization and AI adoption compared to software/finance, with pilots more common than production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-assisted CAD tools substantially augment drafter productivity by automating repetitive geometry, applying standards, and suggesting designs, while the human retains quality control and creative judgment. This is a strong case where AI enhances human capability without necessarily replacing the role. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted drafting tools, auto-dimensioning, and generative layout suggestions meaningfully speed up drafters' workflow while humans retain control over final technical accuracy. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | CAD systems and AI-assisted design tools can automate substantial portions of routine drafting (converting sketches to drawings, applying standards, generating repeated elements), but complex spatial reasoning, judgment about engineering feasibility, and integration of multiple constraints typically require human oversight. The task is partially automatable with current technology. |
| Task automatability | claude-sonnet-5 | 3/5 | CAD/BIM drawing generation from sketches or notes can be partially automated with AI-assisted tools, but translating engineering intent, standards compliance, and complex site-specific detailing still requires substantial human judgment and iteration. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Drafting outputs must often be reviewed and stamped by licensed engineers or architects who bear liability for accuracy and compliance with building codes; this creates a hard requirement for qualified human review. Regulatory frameworks and professional liability standards strongly protect this workflow. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Civil/structural drawings often require professional engineer stamping and regulatory compliance, creating strong liability and licensing barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While CAD automation reduces labor, current systems require significant setup, integration, and human oversight by skilled drafters. The all-in cost (software, training, validation, corrections) remains comparable to or slightly higher than paying drafters for routine work, especially for non-standardized projects. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Current AI tools require significant integration, licensed CAD software, and drafter review, so cost savings are modest rather than order-of-magnitude, especially given liability requirements in construction documents. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed CAD software with parametric modeling and some AI plugins exists in production (e.g., Autodesk, Revit with plugins), but reliable end-to-end conversion from sketches to compliant, production-ready drawings still requires human review and correction. Material error rates and scope limitations persist, particularly for novel or complex structures. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some AI-assisted CAD plugins (e.g., generative design, auto-dimensioning) exist, but no mature product reliably produces full construction-grade drawings for civil infrastructure from notes without heavy drafter oversight. |
Determine the order of work and method of presentation, such as orthographic or isometric drawing.
37CI 30–44 · exposure 33 · augmentation 50 · importance 4.1/5 · click for rater detail
Determine the order of work and method of presentation, such as orthographic or isometric drawing.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architectural and drafting sectors adopt AI-assisted tools at modest pace; while CAD software is ubiquitous, AI-driven workflow optimization and intelligent drawing-method recommendation are still in pilot phases, not mainstream production deployment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC industry is a moderate-to-slow adopter of AI for design workflow decisions, with CAD automation more focused on drawing execution than upfront method selection. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully suggest drawing types and sequence based on project metadata and past projects, speeding up the drafter's decision-making, though human judgment remains essential for final determination of order and presentation method. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can suggest drawing types and organize task order based on similar past projects, giving drafters a useful starting point though final decisions remain human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can assist in generating drawing suggestions and organizing standard presentation sequences (e.g., recommending orthographic projections for technical clarity), but determining the optimal order and method requires domain expertise, project-specific constraints, and judgment about stakeholder needs that currently demand significant human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a planning/judgment step involving project-specific conventions, client needs, and downstream usage; AI can suggest defaults but cannot reliably determine the full workflow and presentation strategy without human oversight.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Professional liability and the requirement that architects/lead drafters sign off on presentation decisions create moderate friction; licensing standards expect human expertise in these determinations, though the barrier is not absolute since the final decision is typically human-approved. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically for this planning task, though drafters often work under engineers/architects who may want final say on presentation approach. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools (e.g., generative drafting assistants) cost roughly equivalent to or sometimes more than the human drafter's time when integration and oversight are factored in, particularly for judgment-heavy decisions requiring domain knowledge. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Human drafters must still make these judgment calls; AI assistance requires setup and review, so cost savings are modest rather than order-of-magnitude. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | CAD and design software exist with some rule-based decision support, but no deployed system reliably determines presentation order and drawing method end-to-end without human review; solutions are narrow (template-based) rather than adaptive to project complexity. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | CAD tools and AI assistants offer templates and drawing-type recommendations, but no deployed product autonomously decides project workflow and presentation method reliably across contexts. |
Analyze building codes, by-laws, space and site requirements, and other technical documents and reports to determine their effect on architectural designs.
36CI 30–43 · exposure 33 · augmentation 63 · importance 4.3/5 · click for rater detail
Analyze building codes, by-laws, space and site requirements, and other technical documents and reports to determine their effect on architectural designs.
36| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architecture and construction are traditionally slower in digital adoption; while some firms pilot AI document review, systematic deployment for code analysis remains limited, and many firms still rely on manual code review by experienced staff. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture/engineering/construction) is a historically slow-adopting, fragmented industry with limited digitization of code interpretation workflows compared to sectors like finance or software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by quickly surfacing relevant code sections, flagging potential conflicts, and summarizing requirements, allowing drafters to focus on design impact assessment and trade-off analysis rather than manual document hunting. This is genuinely useful but not transformative of the full task. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can quickly search, summarize, and cross-reference lengthy code documents and technical reports, substantially speeding up the research phase even though final interpretation remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can extract and summarize text from building codes and technical documents, determining their effect on architectural designs requires nuanced judgment about design implications, trade-offs, and context-specific interpretation that current systems cannot reliably perform end-to-end. Most of the task involves subjective spatial and technical reasoning rather than pure information retrieval. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can extract and cross-reference code requirements against design parameters, but nuanced interpretation of ambiguous local by-laws and site-specific judgment calls still require human review, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Architectural and construction projects typically involve professional liability and regulatory sign-off by licensed architects or engineers; while drafters perform this analysis, errors in code interpretation can create legal exposure, creating some organizational friction but not an absolute barrier to AI use as an assistive tool. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement specifically for this analysis step, but liability for code compliance in final designs typically falls on licensed architects/engineers, creating oversight friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While document processing is cheap, the low accuracy and high need for human verification and re-work mean the effective cost per reliable output remains comparable to or exceeds a drafter's loaded wage for this analytical work. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI tools can cheaply flag potential code issues, but the need for human verification and jurisdiction-specific customization keeps blended costs roughly comparable to a drafter's time for this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No production system reliably performs this end-to-end analysis; AI excels at retrieving and summarizing code snippets but fails at the deeper reasoning about architectural impact and design modification. Experimental tools exist for code search but require substantial human oversight and verification. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some code-compliance-checking products (e.g., automated plan review tools) exist but are narrow in scope, jurisdiction-specific, and not yet reliably deployed across the diversity of codes and site conditions drafters encounter. |
Prepare cost estimates, contracts, bidding documents, and technical reports for specific projects under an architect's or engineer's supervision.
34CI 25–43 · exposure 38 · augmentation 63 · click for rater detail
Prepare cost estimates, contracts, bidding documents, and technical reports for specific projects under an architect's or engineer's supervision.
34| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architecture and engineering firms adopt AI incrementally for administrative tasks but remain cautious with deliverables that carry liability. Adoption is in early pilot phase, with conservative organizational cultures slowing deployment into production workflows. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture/engineering/construction) sectors have historically been slower adopters of AI compared to finance or professional services, though drafting and documentation tools are seeing growing pilot use. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist drafters by generating first drafts of boilerplate text, organizing information into report templates, and checking for completeness, raising productivity on routine document assembly while the drafter handles calculations and judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully speed up drafting of technical reports, cost estimate templates, and contract language, letting drafters focus on verification and project-specific judgment while the human remains in control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with document generation and boilerplate contract language, but cost estimates and technical reports require project-specific calculations, site conditions, and professional judgment that current systems cannot reliably produce end-to-end. Human review and modification would be extensive. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft cost estimates, contract boilerplate, and technical report sections quickly, but integrating accurate project-specific quantities, current pricing, and jurisdiction-specific bidding requirements still requires substantial human verification and setup.rating reflects partial but not full automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | These documents carry legal and financial liability; architects and engineers typically must sign off on estimates and reports. Regulatory and professional responsibility requirements mean AI output cannot bypass human review, creating a hard adoption barrier. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Cost estimates, contracts, and bidding documents typically require sign-off and supervision by a licensed architect or engineer, with legal and liability implications that limit unsupervised AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for document drafting (GPT, specialized legal/technical templates) are cheap, but the oversight burden—verifying numbers, checking calculations, ensuring legal compliance—means total cost per task remains substantial relative to a drafter's wage. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cut drafting time for reports and contract templates substantially, but the need for professional review, data verification, and liability sign-off keeps overall cost savings moderate rather than order-of-magnitude. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can draft templates and generate text, deployed products lack the domain-specific accuracy needed for binding cost estimates and contracts without significant human rework. No production system reliably handles the project-specific technical depth required. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Estimating software with AI features and LLM-based document drafting tools exist and are used in practice, but reliable end-to-end generation of accurate cost estimates and bidding documents without heavy human correction is not yet standard in production. |
Determine procedures and instructions to be followed, according to design specifications and quantity of required materials.
32CI 25–39 · exposure 33 · augmentation 63 · importance 3.8/5 · click for rater detail
Determine procedures and instructions to be followed, according to design specifications and quantity of required materials.
32| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architectural and construction firms have adopted CAD and computational tools slowly relative to software-native industries; AI-driven procedure generation remains in pilot phases rather than production deployment at scale in most firms. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Architecture and construction sectors are slower adopters of AI compared to finance or software, with pilots for design assistance emerging but production-scale autonomous procedure-setting rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist drafters by automating routine calculations, suggesting material lists, and generating draft procedures that humans then review and refine, improving productivity on the specification-to-procedure workflow without replacing human oversight. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can help drafters quickly reference standards, estimate quantities, and draft procedural checklists, meaningfully speeding up parts of this task while the drafter retains final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can assist in interpreting design specifications and generating material quantity procedures, but the task requires design judgment, error verification, and integration with project-specific constraints that still need human oversight. Partial automation of procedure generation is feasible, but full end-to-end execution with 50% time savings and equal quality is not yet demonstrated. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires integrating design intent, code compliance, and material logistics into procedural judgment calls that current AI cannot reliably perform end-to-end without heavy human oversight.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Design specifications and material procedures often require sign-off by licensed architects or engineers, and errors can lead to costly construction failures, creating both regulatory and liability barriers that require human professional judgment and accountability. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While no strict licensing requirement exists for drafters, liability for construction errors and reliance on human judgment in interpreting specs creates meaningful organizational friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools for this task require significant human oversight, verification, and refinement, making the all-in cost (inference plus integration plus quality control) comparable to or higher than a human drafter performing the work, especially given liability concerns. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Given the need for extensive human verification and correction of AI-generated procedural decisions, current AI assistance does not yet yield large net cost savings over skilled drafters. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While CAD software can assist with calculations and some specification interpretation, no deployed product reliably generates complete and legally-sound procedures and instructions from design specs autonomously. Products exist for material takeoffs and some procedural suggestions, but material error rates in complex projects remain significant. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some CAD-integrated tools can suggest specifications or quantities, but no deployed product autonomously determines full procedures and instructions reliably across varied projects. |
Review rough sketches, drawings, specifications, and other engineering data to ensure that they conform to design concepts.
30CI 30–30 · exposure 25 · augmentation 50 · click for rater detail
Review rough sketches, drawings, specifications, and other engineering data to ensure that they conform to design concepts.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architecture and civil engineering firms have digitized slowly relative to software and finance. Pilot AI tools exist, but production adoption remains limited; most firms still rely on human drafters and senior engineers for review, with AI assistance not yet mainstream. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture/engineering/construction) sectors are historically slow digitizers with fragmented software ecosystems, though BIM-adjacent AI tools are emerging in pilots. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist drafters by automatically flagging dimensional mismatches, missing annotations, and common standard deviations, allowing humans to focus on higher-level design logic and intent. This assistive capability meaningfully raises productivity on parts of the review task while the drafter remains in control. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by flagging obvious inconsistencies, dimension errors, or code-compliance issues, speeding up parts of the review while the drafter retains final judgment on design intent conformance. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can identify some visual discrepancies and flag obvious deviations from stated specifications, but engineering design review requires contextual judgment about feasibility, buildability, code compliance, and design intent that remains difficult for current systems. Full end-to-end automation with 50% time savings at equal quality is not demonstrable. |
| Task automatability | claude-sonnet-5 | 2/5 | Reviewing sketches against design concepts and specifications requires visual-spatial judgment and contextual understanding of engineering intent that current AI handles only partially and unreliably at scale. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Design review carries liability risk—errors can cascade into costly field revisions or safety issues—creating organizational and reputational friction against full automation. However, there is no hard legal licensing requirement mandating human sign-off in all jurisdictions, leaving moderate barriers. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement for drafters specifically, but liability for design conformance and downstream engineering/construction errors creates organizational caution about full automation of this checking function. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools for design review require significant setup, integration with CAD systems, prompt engineering, and human oversight to catch errors. The total cost per review task remains comparable to or higher than hiring a junior drafter for the same output quality. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI review tools require significant setup, integration with CAD systems, and human verification, making costs comparable to or only modestly cheaper than a drafter's review time for this judgment-heavy task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Vision models can detect geometric inconsistencies and perform basic compliance checks, but deployed products lack the nuanced understanding of engineering standards, interdisciplinary constraints, and design tradeoffs needed for reliable production review. Narrow demos exist; production reliability at scale is absent. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some CAD-integrated AI tools flag inconsistencies or code violations, but no deployed product reliably reviews rough sketches against design intent across the diversity of architectural/civil projects. |
Correlate, interpret, and modify data obtained from topographical surveys, well logs, and geophysical prospecting reports.
30CI 30–30 · exposure 25 · augmentation 63 · importance 3.9/5 · click for rater detail
Correlate, interpret, and modify data obtained from topographical surveys, well logs, and geophysical prospecting reports.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architectural and civil engineering sectors have adopted digital tools broadly, but adoption of AI-driven geophysical data interpretation remains pilot-stage and concentrated in large firms; most small-to-mid practices still rely on manual methods. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Civil/architectural drafting and geotechnical fields have moderate digitization but remain slower to adopt AI agents compared to information-sector white-collar work, with most tools still supplementary. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by automating data extraction, flagging anomalies, and generating initial correlation hypotheses, but the drafter must validate interpretations and make final modifications, making this a productivity aid rather than a transformative replacement. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted GIS platforms, data visualization, and pattern recognition tools can meaningfully speed up correlation and preliminary interpretation of survey data, keeping the drafter in the loop for validation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can parse and extract structured data from surveys and logs, correlating disparate geophysical datasets and making informed modifications requires domain expertise and judgment about subsurface conditions that current systems cannot reliably perform end-to-end. The interpretive synthesis across multiple data sources remains largely manual. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires integrating heterogeneous spatial and geological data with domain judgment about terrain, subsurface conditions, and engineering implications, which current AI cannot reliably do end-to-end without heavy human oversight.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Professional liability concerns and regulatory requirements in civil/architectural projects create friction; however, no hard licensing bar explicitly prohibits AI from performing data correlation, though sign-off by a licensed professional is often required. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While no strict licensing mandates a drafter perform this specific task, professional oversight (engineers, surveyors) and liability for interpretation errors create meaningful friction against pure automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for geophysical analysis require significant specialized training, domain licensing, and oversight integration; total cost per usable output often remains comparable to or exceeds the cost of a skilled drafter performing the work directly. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized software and domain expertise are still required for interpretation, so AI reduces some labor but does not yet approach order-of-magnitude cost savings given verification needs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Limited deployed products perform this specific correlation task reliably in production; most applications are narrow (e.g., well-log interpretation in specific geologies) or require heavy human validation and rework. General-purpose geophysical data synthesis lacks production maturity. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some GIS and geospatial AI tools assist with data overlay and pattern detection, but no deployed product autonomously correlates and interprets topographic, well-log, and geophysical data reliably in production. |
Locate and identify symbols on topographical surveys to denote geological and geophysical formations or oil field installations.
30CI 25–35 · exposure 25 · augmentation 50 · importance 3.5/5 · click for rater detail
Locate and identify symbols on topographical surveys to denote geological and geophysical formations or oil field installations.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architectural and civil drafting remains concentrated in small to mid-sized firms with legacy workflows and low digitization; adoption of specialized AI for survey symbol identification is minimal and largely confined to large engineering firms running pilots. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Architectural/civil drafting and oil field surveying are moderately digitized but adoption of AI-driven symbol recognition in this niche is still nascent, mostly pilot projects rather than widespread deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted symbol highlighting and preliminary classification could accelerate drafter review cycles, reducing fatigue-driven errors and speeding up survey interpretation, though human validation remains essential. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-assisted image recognition and CAD software can help drafters quickly locate and classify common symbols, improving speed while the drafter still verifies and interprets results. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While computer vision can detect and classify symbols on scanned topographical surveys, the task requires contextual interpretation of geological formations and field installations that demands domain expertise. Current AI systems struggle with the nuanced spatial reasoning and rare symbol variants that expert drafters routinely handle. |
| Task automatability | claude-sonnet-5 | 2/5 | Symbol identification on topographical surveys requires domain-specific visual interpretation with high accuracy needs; current AI can assist but not reliably replace this end-to-end without significant human verification.", "partial computer vision tools exist but are narrow and unproven at scale for this specific task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Professional liability for incorrect geological or oil field installation identification creates high error-cost asymmetry; moreover, professional engineering standards and regulatory compliance for survey documentation often require human expert review and sign-off. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically restricts this task to a human, though industry-specific accuracy standards and downstream engineering reliance create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI symbol recognition services require significant custom training, manual validation, and domain expert oversight, making the all-in cost per survey comparable to or exceeding a drafter's hourly rate for this specialized task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools require specialized training data and integration with survey/CAD systems, and human verification is still needed for accuracy, keeping costs comparable to skilled drafter labor rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Document analysis and symbol recognition products exist, but none are purpose-built and validated for specialized geological/oil-field survey symbol identification at production scale. Prototype systems show promise but lack the reliability and accuracy required for professional drafting work. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some GIS and CAD-integrated tools offer symbol recognition, but few production systems reliably automate full identification of geological/geophysical symbols on topographic surveys without human review. |
Create freehand drawings and lettering to accompany drawings.
30CI 25–35 · exposure 20 · augmentation 63 · importance 3.5/5 · click for rater detail
Create freehand drawings and lettering to accompany drawings.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architectural and drafting firms operate in a design-focused, human-collaborative environment with strong aesthetic and technical standards; adoption of AI for freehand work remains limited and is largely experimental rather than production-normalized. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Architecture and drafting are moderately digitized but physical/creative sketching tasks lag behind digital-only tasks in AI tool adoption; most firms use AI for other design/documentation tasks, not freehand sketching. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can significantly assist drafters by generating initial sketches, automating lettering placement and formatting, or proposing design variations, allowing human drafters to iterate faster and focus on refinement and judgment rather than time-consuming manual work. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI sketch-generation and style-transfer tools can assist by generating rough concept visuals or providing inspiration/backgrounds, moderately aiding ideation, but do not replace the drafter's own freehand execution. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Freehand drawing requires artistic judgment, spatial reasoning, and aesthetic decision-making that current AI systems struggle with authentically. While AI can generate drawings or add lettering, reproducing the specific style, proportion, and expressive quality of human freehand work—especially when it must integrate with existing technical drawings—falls short of 50% time savings at equal quality for most real-world scenarios. |
| Task automatability | claude-sonnet-5 | 2/5 | Freehand sketching and lettering are manual, tactile skills tied to hand-eye coordination and rapid conceptual expression; AI image generation can produce sketch-like outputs but cannot replicate the interactive, context-specific freehand drafting workflow at equal quality end-to-end.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Professional standards and client expectations for authentic freehand work create moderate friction; there is no hard legal barrier, but architectural firms often value human-created aesthetics and accept liability for drawing accuracy, which slows automation adoption. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing barrier specifically protects freehand sketching itself, though drafters' work is often reviewed by licensed architects/engineers, creating mild organizational oversight friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI image generation and lettering tools are relatively cheap per inference, but integration costs, prompt engineering, revision cycles, and quality control oversight make the all-in cost competitive with or potentially higher than a drafter's labor for final-quality output. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI image generation is cheap per output, it cannot substitute for the specific task's integration into professional documentation, so effective cost-per-equivalent-output remains high once rework and validation are factored in. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No production systems reliably generate freehand drawings and lettering that meet professional architectural standards and integrate seamlessly with technical documentation. AI tools can assist with lettering or generate stylized sketches, but deploying them as substitutes for this task in professional practice remains rare and inconsistent. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs freehand architectural sketching and hand-lettering as part of a drafter's workflow; generative AI tools produce static images, not integrated freehand drawings tied to project documentation. |
Coordinate structural, electrical, and mechanical designs and determine a method of presentation to graphically represent building plans.
29CI 25–32 · exposure 25 · augmentation 75 · importance 4.4/5 · click for rater detail
Coordinate structural, electrical, and mechanical designs and determine a method of presentation to graphically represent building plans.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Architecture and engineering firms are experimenting with AI-assisted CAD and generative tools, but adoption remains in pilot phases; few have moved to production-scale autonomous coordination, given risk and regulatory constraints. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC (architecture, engineering, construction) is a traditionally slow-adopting sector for AI compared to information/finance industries, though BIM and CAD software adoption is fairly mature and evolving with AI plugins. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist drafters by suggesting design alternatives, automating layout variants, flagging conflicts between disciplines, and accelerating presentation rendering—activities that strengthen human productivity while preserving expert judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enhanced CAD/BIM tools significantly help drafters detect clashes, generate draft layouts, and visualize presentation options, meaningfully boosting productivity while humans retain final coordination responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with individual design components and suggest presentation methods, coordinating multiple engineering disciplines requires substantial human judgment about trade-offs, conflicts, and buildability. Current AI tools cannot reliably resolve competing design constraints or make authoritative coordination decisions end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Coordinating multi-discipline design clashes and choosing presentation methods requires integrative judgment across trades that current AI cannot reliably perform end-to-end, though it can assist with clash detection and drafting sub-tasks. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Building codes, professional licensing (PE/architect sign-off), and liability for structural/safety decisions create regulatory and legal barriers. Plans typically require licensed professional review and signature, preventing full autonomous substitution. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for drafters themselves, but liability for coordination errors in construction documents and organizational reliance on experienced staff create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration of AI into design workflows still requires specialized personnel and substantial oversight; the human drafter cost remains competitive with AI infrastructure and human validation overhead for coordination-level work. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can reduce some drafting time but human oversight, coordination meetings, and judgment calls on presentation remain necessary, keeping costs closer to comparable rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | CAD software and generative design tools exist but require heavy human steering; no deployed AI system reliably coordinates multi-discipline conflicts or generates presentation-ready building plans autonomously. Most production use remains in narrow, well-constrained sub-tasks. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | BIM software has automated clash-detection features and some AI-assisted drafting tools exist, but full cross-discipline coordination and presentation-method decisions are still handled by trained drafters in production workflows. |
Explain drawings to production or construction teams and provide adjustments as necessary.
28CI 25–30 · exposure 25 · augmentation 63 · importance 3.6/5 · click for rater detail
Explain drawings to production or construction teams and provide adjustments as necessary.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Construction and drafting sectors are relatively laggard in AI adoption compared to information and finance; production teams remain geographically distributed, often on-site, and dependent on synchronous human communication. Adoption remains in pilot phases rather than production deployment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Construction and architecture remain lower-digitization, physically grounded sectors with slower AI adoption for on-site coordination tasks compared to purely digital information work. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist drafters by generating initial explanatory text, suggesting clarifications, or producing alternative interpretations of drawings. However, the human drafter must validate and adapt explanations for site conditions, making this a genuinely assistive rather than transformative tool. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist drafters by generating clearer visualizations, auto-annotating changes, and drafting explanatory notes or FAQs ahead of team discussions, boosting communication efficiency while the human remains central. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate text explanations of drawings and suggest minor adjustments, explaining technical drawings to construction teams requires real-time clarification, site-specific problem-solving, and the ability to negotiate competing constraints—tasks that demand human judgment and accountability. Current AI cannot reliably handle the iterative, context-dependent nature of this task. |
| Task automatability | claude-sonnet-5 | 2/5 | This task involves live, interactive communication and real-time adjustment based on questions and site context, which current AI cannot fully replace end-to-end despite being able to summarize or annotate drawings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Legal and professional liability is substantial: miscommunication in construction can cause safety failures, cost overruns, and injury. Professional liability insurance, contractual responsibility, and industry standards typically require a licensed or certified professional to be accountable for explanations provided to production teams. |
| Adoption barriers | claude-sonnet-5 | 3/5 | There's no strict licensing requirement for this explanatory task, but organizational reliance on human judgment, liability for construction errors, and the need for real-time interactive problem-solving create moderate friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The oversight and human verification needed to validate AI explanations would approach or exceed the cost of having a drafter do the task directly. Integration and liability management add significant overhead, making this more expensive than direct human labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While document Q&A tools are cheap, the human-in-the-loop coordination, trust-building, and on-site clarification still require a paid drafter or engineer, keeping AI's incremental cost savings limited for this specific interactive task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably explains complex architectural drawings to construction teams with the flexibility and accountability required in production settings. While AI can draft explanatory text, it lacks the ability to understand site conditions, respond to unexpected questions, and validate that teams have correctly interpreted critical details. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools can generate drawing annotations or answer some technical queries about CAD files, but no deployed product reliably substitutes for a drafter explaining plans to a construction team and negotiating adjustments in real-world settings. |
Plot characteristics of boreholes for oil and gas wells from photographic subsurface survey recordings and other data, representing depth, degree, and direction of inclination.
28CI 25–30 · exposure 25 · augmentation 50 · importance 3.1/5 · click for rater detail
Plot characteristics of boreholes for oil and gas wells from photographic subsurface survey recordings and other data, representing depth, degree, and direction of inclination.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | The oil and gas sector digitizes slowly compared to software/finance; legacy systems dominate, and many firms operate with established manual or semi-manual drafting workflows. Pilot automation projects exist but production deployment of end-to-end borehole plotting remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Oil and gas sector adoption of AI is growing but remains slower than software/finance sectors, especially for specialized technical drafting functions tied to legacy engineering workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by auto-extracting and pre-processing borehole data, suggesting plot layouts, and flagging anomalies, allowing a drafter to review and refine rather than manually transcribe all measurements. This augmentation is real but incremental, not transformative. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI and specialized software can assist drafters by automating calculations, flagging anomalies, and generating preliminary plots, improving efficiency while humans verify accuracy against critical borehole data. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can extract numeric data from images and plot coordinates, the task requires integrating multiple data modalities (photographic surveys, depth readings, inclination angles) and producing precise technical plots that must meet engineering standards. Current vision+plotting systems lack the domain-specific validation and error-checking that would achieve 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Interpreting subsurface survey recordings requires specialized geological/directional survey judgment and integration with well planning that current general AI tools cannot reliably automate end-to-end, though data plotting from structured logs is partially automatable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Oil and gas well documentation is subject to regulatory requirements and industry standards (API, SPE) that mandate professional review and sign-off. Liability for incorrect borehole characterization is high, and wells are safety-critical infrastructure, creating strong regulatory and professional barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not requiring a professional license akin to a PE stamp for drafting itself, the work feeds into safety-critical well planning where errors have high consequences, creating organizational and liability-driven caution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI vision and plotting inference costs are low, but integration, domain-specific model training, and required human oversight to verify subsurface survey data interpretation add significant overhead. The all-in cost remains comparable to or exceeds a drafter's hourly rate for this specialized task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized survey-processing software and domain expertise still require significant licensing, integration, and skilled oversight, keeping costs closer to human-comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | General image analysis and data extraction tools exist, but no mature production system specifically handles borehole survey interpretation and automated technical plotting with the accuracy required for oil/gas engineering. Specialized oil/gas software exists but typically requires manual data entry rather than end-to-end automation. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some specialized oilfield software automates directional survey calculations and plotting, but these are narrow, vendor-specific engineering tools rather than general AI products demonstrating broad reliable performance on this niche drafting task. |
Obtain and assemble data to complete architectural designs, visiting job sites to compile measurements as necessary.
23CI 21–25 · exposure 16 · augmentation 50 · importance 3.5/5 · click for rater detail
Obtain and assemble data to complete architectural designs, visiting job sites to compile measurements as necessary.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Architecture and construction remain relatively lower-adoption sectors for full automation due to the physical and spatial nature of the work. While CAD and 3D tools are standard, autonomous site measurement and data compilation remain niche, with most firms still relying on manual site surveys. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Architecture and drafting fields are moderate adopters of AI for design assistance, but the physical site-visit component is inherently resistant to digital automation and sees minimal AI penetration. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by organizing measurement data, suggesting measurement points, cross-referencing to code, and automating documentation once raw data is collected. However, the human must still visit the site and decide what to measure, limiting transformative impact. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI tools like LiDAR scanning apps, photogrammetry, and drafting software can help process and organize field-collected data into designs, improving productivity for the data-assembly portion of the task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with data assembly and organization tasks remotely, visiting job sites to compile measurements requires physical presence and real-time environmental assessment that current AI systems cannot perform. Only the remote data compilation portion could be meaningfully automated, leaving the on-site measurement collection entirely dependent on human labor. |
| Task automatability | claude-sonnet-5 | 2/5 | The physical site visit and measurement collection require human presence and manual data-gathering; AI cannot independently visit sites or take physical measurements, though it can assist with data assembly afterward.5 |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Professional licensure (PE/architect oversight) and liability for design accuracy create moderate friction; incorrect measurements can lead to costly errors. Customer preference and legal responsibility for on-site assessment keep humans in the loop, though supervisory automation is plausible. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No formal licensing barrier for measurement-taking itself, but physical presence requirement and reliance on human judgment for site conditions create practical friction against remote automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Remote data processing tools can reduce some overhead, but the necessity of site visits by trained professionals (who understand what to measure and why) keeps overall costs high relative to AI-only approaches. The human cost of the site visit dominates the economics. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since the core site-visit and measurement work cannot be automated, human labor cost remains necessary regardless of AI tools used for data assembly. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably visits physical job sites, takes accurate architectural measurements, and assembles them into design-ready data autonomously. Current photogrammetry and LiDAR tools exist but require significant human direction and interpretation; they are not end-to-end solutions in production architectural workflows. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously visits job sites and collects physical measurements; this remains a human-performed field task with no production AI substitute. |
Supervise or conduct field surveys, inspections, or technical investigations to obtain data required to revise construction drawings.
19CI 7–30 · exposure 13 · augmentation 63 · importance 3.7/5 · click for rater detail
Supervise or conduct field surveys, inspections, or technical investigations to obtain data required to revise construction drawings.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While construction and civil engineering sectors show growing interest in drone and LiDAR adoption, actual deep integration into standard practice remains uneven; many firms still rely primarily on traditional field personnel, and regulatory/liability concerns slow rapid substitution. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Construction and civil engineering sectors are historically slow adopters of AI for physical field tasks, with digitization concentrated in office-based drafting rather than site inspection. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Drone imagery, automated point-cloud analysis, real-time 3D site visualization, and AI-assisted anomaly detection substantially assist field technicians in locating and documenting issues faster and more comprehensively than manual inspection alone, while the human expert remains responsible for interpretation and sign-off. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI tools (drone imagery analysis, automated point-cloud processing, image-based defect detection) can help process and interpret field data faster once collected, meaningfully aiding the analysis portion of this task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While LiDAR scanning, drone imagery, and AI-powered site analysis can capture much field data, current systems cannot fully replace the judgment-intensive aspects of conducting comprehensive inspections—evaluating structural conditions, assessing site accessibility, or identifying hazards—at the quality and thoroughness expected, particularly for complex projects. Autonomous field surveying exists but requires human verification and interpretation. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical presence on-site, operating survey equipment, and visually inspecting construction conditions—none of which current AI systems can perform end-to-end without a human physically executing the fieldwork. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Field surveys and inspections for construction projects are often governed by professional engineering licensure requirements and liability standards; many jurisdictions legally require a licensed professional engineer or surveyor to certify findings, creating a hard barrier to full automation and human substitution. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing barrier for drafters doing field surveys, but physical site access, safety requirements, and liability for inspection accuracy create real friction against remote/AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Drone and scanning equipment with associated software involves significant upfront capital and licensing costs, plus integration overhead; for many small-to-mid-sized civil projects, the per-task cost remains higher than deploying experienced field technicians, particularly when accounting for equipment downtime and interpretation labor. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since the physical fieldwork must be done by a person (or with equipment operated/supervised by one), AI does not replace the core labor cost; any AI cost is additive, not substitutive. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Automated or semi-automated field inspection tools (drones, LiDAR) exist but are still constrained by weather, site complexity, and the need for human technicians to operate equipment and validate findings; AI-powered site analysis remains in the pilot/narrow-scope phase rather than mature production deployment across civil projects. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously conducts field surveys or physical inspections; AI-assisted photogrammetry/LiDAR analysis exists but still requires human data capture and supervision on-site. |
Represent architect or engineer on construction site, ensuring builder compliance with design specifications and advising on design corrections, under supervision.
5CI 5–5 · exposure 0 · augmentation 38 · click for rater detail
Represent architect or engineer on construction site, ensuring builder compliance with design specifications and advising on design corrections, under supervision.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Construction and site supervision remain highly physical, regulatory-bound, and reliant on human judgment and presence; the sector has low adoption velocity for agent-based automation of compliance representation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Construction site oversight is a physical, low-digitization activity in a sector with historically slow AI adoption for hands-on field roles. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with visual defect detection or document cross-referencing, but the core task—representing the architect, advising on corrections, and ensuring compliance—fundamentally requires human authority and presence, limiting augmentation value. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with reviewing specifications, flagging discrepancies from photos or documents, and generating reports, but the on-site advisory and verification interaction itself remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time on-site presence, visual inspection of complex construction details, interpersonal interaction with builders, and judgment calls on design deviations—capabilities far beyond current AI systems. No end-to-end automation is feasible today. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical presence on-site, real-time judgment, and interaction with builders to verify compliance and resolve issues, none of which current AI systems can perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: construction compliance and sign-off often require licensed professionals (PE/RA in many jurisdictions), liability attaches to the person certifying compliance, and design corrections demand professional judgment and accountability that cannot be delegated to an unattended AI system. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Site representation often involves professional liability, contractual authority, and safety compliance responsibilities that typically require a qualified human presence, even if under supervision rather than full licensure. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Current AI systems cannot replace the core function. Any attempted automation would require human oversight, site visits, and fallback to human judgment, making the total cost comparable to or higher than a human drafter or site representative. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this on-site task, so cost comparison favors the human by default since AI cannot deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs this task reliably. While computer vision can detect some defects in images, the integration with site coordination, decision-making, and advisory authority under supervision remains entirely in the research/demo stage. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product acts as an on-site representative for construction compliance oversight; this remains a physical, in-person supervisory role. |
Supervise and train other technologists, technicians, and drafters.
4CI 0–7 · exposure 0 · augmentation 38 · importance 4.1/5 · click for rater detail
Supervise and train other technologists, technicians, and drafters.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Organizations have shown minimal appetite to automate supervisory functions; human managers remain the standard practice across all sectors due to legal requirements, organizational culture, and the need for discretionary judgment in personnel matters. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | AEC firms are moderate adopters of AI tools for design tasks, but supervisory/managerial functions see little to no AI displacement. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can provide some supplementary assistance (e.g., generating training materials, tracking employee progress, scheduling), but cannot meaningfully augment the core supervisory and mentoring functions that require human presence, judgment, and relationship-building. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help create training materials, standardize onboarding documentation, and answer technical questions, aiding but not replacing supervisory activity. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Supervision and training of technicians requires real-time interpersonal engagement, performance assessment, mentoring, and adaptive feedback tailored to individual learners. Current AI systems cannot reliably conduct the interactive, contextual, and emotionally-informed management that this task demands. |
| Task automatability | claude-sonnet-5 | 1/5 | Supervising and training people is an interpersonal, leadership task requiring mentorship, judgment, and real-time feedback that current AI cannot perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Supervisory and training authority typically require a human in a formal management position by organizational policy and law (labor law, duty of care). Responsibility for employee development, performance evaluation, and conduct cannot be legally or ethically delegated to an automated system without human sign-off and accountability. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Organizational structure, accountability, and liability for staff performance require a human supervisor; delegating supervision to AI faces strong institutional and legal friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Substituting human supervisors with AI to train and oversee technologists would be prohibitively expensive relative to deploying experienced human supervisors, given the liability, accountability, and the need for human judgment in staff development and discipline. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for the supervisory function itself, so cost comparison favors the human role heavily. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products perform supervisory and training responsibilities autonomously in professional environments. While AI can assist with training content creation or documentation, the actual supervision of personnel—evaluating competence, providing corrective feedback, and career development—remains entirely human-dependent in practice. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product manages or trains human technical staff autonomously; AI is at best a supplementary training-content tool. |
Related occupations — Architecture & Engineering
How to read this
A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.
What would change this score
New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.