Civil Engineers

17-2051.00

Median wage $100,840/yr367,840 employed (US)Rank #415 of 923 scored · top 45% by substitution

Perform engineering duties in planning, designing, and overseeing construction and maintenance of building structures and facilities, such as roads, railroads, airports, bridges, harbors, channels, dams, irrigation projects, pipelines, power plants, and water and sewage systems.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution29

Exposure30

Augmentation68

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

16 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

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%31

panel mean rating 2.2/5 → substitution pressure 31/100

Technical feasibility todayw 20%30

panel mean rating 2.2/5 → substitution pressure 30/100

Cost vs. human wagew 15%31

panel mean rating 2.3/5 → substitution pressure 31/100

Adoption barriersw 20%inverted — strong barriers lower the score24

panel mean rating 4.1/5 (barrier strength) → substitution pressure 24/100

Sector adoption velocityw 10%31

panel mean rating 2.2/5 → substitution pressure 31/100

Task breakdown (16 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.

Estimate quantities and cost of materials, equipment, or labor to determine project feasibility.

CI 50–75 · exposure 62 · augmentation 88 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Large engineering and construction firms have integrated AI cost-estimation tools into standard workflows over the past 5–10 years. Smaller firms adopt more slowly, but the sector overall shows strong, measurable adoption of these technologies in competitive bidding and preliminary design phases.
Sector adoption velocity	claude-sonnet-5	3/5	Construction and civil engineering sectors are historically slow tech adopters, though BIM-integrated estimating tools and preconstruction software are increasingly used, placing this at middling adoption speed.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI dramatically amplifies estimator productivity by automating quantity extraction, cost lookups, and scenario modeling, while the engineer retains judgment on feasibility, risk adjustment, and client-specific factors. This is a textbook case of human-AI partnership raising output quality and speed.
Augmentation potential	claude-sonnet-5	4/5	AI-driven quantity takeoff, cost databases, and predictive analytics substantially speed up the estimation process while engineers retain responsibility for final judgment and validation.
Task automatability	claude-haiku-4-5-20251001	4/5	AI can reliably extract and aggregate material quantities from specifications, cross-reference unit costs from databases, and calculate labor estimates based on standard crews and productivity rates. Most of the task—parsing, lookup, arithmetic, and feasibility threshold checks—is automatable; human judgment on site-specific risk factors and contingencies remains, but the time-saving threshold is met.
Task automatability	claude-sonnet-5	3/5	AI tools can pull quantities from BIM models and generate preliminary cost estimates, but final feasibility estimates require judgment on site conditions, risk contingencies, and local market pricing that current systems only partially capture.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	Estimation itself carries no licensing mandate—any company can automate it. However, regulatory and contractual expectations that a licensed civil engineer or architect sign off on feasibility findings create soft friction. Liability concerns (if an underestimate causes budget overruns) may slow adoption, but do not legally require human performance.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement mandates a PE sign this specific estimate, but liability for inaccurate feasibility studies and organizational reliance on experienced engineers create moderate friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI-powered estimation tools cost thousands to tens of thousands annually (subscriptions, data feeds) versus a senior estimator's $100k+ loaded wage. For organizations processing many projects, the per-estimate cost is far lower; even for single projects, AI overhead amortizes favorably over cycles.
Cost vs. human wage	claude-sonnet-5	3/5	AI-assisted takeoff and estimation tools reduce hours spent on quantity surveying, but human oversight, data integration, and validation still require significant engineering time, keeping costs roughly comparable rather than order-of-magnitude cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Construction estimation software with AI modules (cost databases, quantity takeoff automation, labor rate calculators) is widely deployed in professional practice. Reliability is high for routine projects; edge cases and complex/novel designs still benefit from expert review, but core estimation workflows are production-grade.
Technical feasibility today	claude-sonnet-5	3/5	Cost-estimation software with AI features (e.g., automated takeoff tools) is deployed in industry, but engineers still heavily verify and adjust outputs due to material/labor price volatility and project-specific nuances.

Compute load and grade requirements, water flow rates, or material stress factors to determine design specifications.

CI 45–69 · exposure 62 · augmentation 100 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Civil engineering firms, especially larger practices and infrastructure consultancies, have been rapidly adopting AI-assisted design and computational tools over the past 5–10 years. BIM integration, computational design platforms, and AI-enhanced FEA are now mainstream in professional practice, reflecting high adoption depth in the sector.
Sector adoption velocity	claude-sonnet-5	3/5	Engineering firms increasingly use AI-assisted analysis and generative design tools, but adoption is moderate and uneven compared to fast-moving digital-native sectors, with pilots more common than full production reliance.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI dramatically augments engineer productivity on this task by automating iterative load calculations, instantly re-computing stress factors for design variants, and suggesting optimized material specifications. Engineers remain in the loop for judgment and sign-off while AI handles the computational heavy lifting, transforming design iteration speed and exploring more design alternatives.
Augmentation potential	claude-sonnet-5	5/5	AI and computational tools dramatically speed up load, flow, and stress calculations, letting engineers iterate designs faster while retaining responsibility for final judgment and compliance.
Task automatability	claude-haiku-4-5-20251001	4/5	Current AI systems, including engineering software with embedded machine learning and AI-assisted computational tools, can reliably compute load calculations, stress factors, and water flow rates based on standard formulas and established engineering principles. While final design decisions require human judgment, the computational core—determining specifications from inputs—is highly automatable and achieves >50% time savings with quality parity.
Task automatability	claude-sonnet-5	3/5	AI/engineering software can perform many standardized calculations (load, flow, stress) quickly, but final design specification requires professional judgment, code compliance checks, and site-specific integration that current AI cannot fully own end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	A licensed Professional Engineer (PE) must typically review, certify, and seal structural calculations and design specifications in most jurisdictions, creating a hard legal barrier to full automation. The PE's liability signature cannot be delegated to AI, mandating human sign-off on final specifications.
Adoption barriers	claude-sonnet-5	4/5	Civil engineering design specifications typically require a licensed Professional Engineer's stamp/sign-off due to public safety and liability regulations, creating a strong legal barrier to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Once licensed engineering software is in place, computational inference cost is negligible compared to the loaded hourly wage of a licensed civil engineer ($75–120/hour). A computation that might take an engineer 1–2 hours costs pennies in software licensing and cloud compute.
Cost vs. human wage	claude-sonnet-5	3/5	Specialized engineering software and cloud compute reduce calculation time significantly, but licensing, validation, and required professional review keep overall cost comparable to, not order-of-magnitude cheaper than, human-performed analysis with tools.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature engineering software packages (ANSYS, SAP2000, HEC-RAS, and others) routinely perform these calculations in production environments across the civil engineering industry. AI-augmented design tools and finite element analysis systems are deployed at scale with documented reliability, though occasional human review of edge cases remains standard practice.
Technical feasibility today	claude-sonnet-5	3/5	Structural/civil analysis software (e.g., SAP2000, HEC-RAS, AI-assisted CAD plugins) reliably performs sub-calculations in production, but full autonomous determination of design specs without engineer oversight is not deployed at scale.

Analyze survey reports, maps, drawings, blueprints, aerial photography, or other topographical or geologic data.

CI 43–70 · exposure 62 · augmentation 88 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Civil engineering and construction firms, particularly in developed markets and large organizations, have rapidly adopted GIS, drone imaging, and AI-assisted document analysis over the past 3–5 years. Pilots and production deployments are now common in major firms and government agencies.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction sectors are historically slow AI adopters relative to information/finance industries, with pilots more common than widespread production deployment for this specific task.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI dramatically amplifies engineer productivity by rapidly extracting, organizing, and highlighting patterns in large volumes of surveys, maps, and aerial data, freeing engineers to focus on interpretation, trade-off analysis, and decision-making. This is a textbook case of human-AI augmentation in technical work.
Augmentation potential	claude-sonnet-5	4/5	AI tools meaningfully speed up preliminary data review, pattern detection in imagery, and drafting of analysis summaries, letting engineers focus judgment on interpretation and decision-making.
Task automatability	claude-haiku-4-5-20251001	4/5	AI can automatically extract, classify, and summarize data from maps, blueprints, aerial photography, and survey reports with high accuracy using computer vision and document parsing. However, the task still requires human judgment on complex interpretations, site-specific constraints, and integration with project context, preventing a full 5 rating despite achieving >50% time savings.
Task automatability	claude-sonnet-5	3/5	AI can extract features from aerial imagery and parse structured drawings/data, but integrating heterogeneous survey/geologic sources into engineering-grade analysis still requires substantial human judgment and setup, limiting full end-to-end automation today.
Adoption barriers	claude-haiku-4-5-20251001	3/5	While analysis and data extraction can be largely automated, civil engineering projects often require a licensed engineer's final sign-off on design decisions and liability sits with the professional. This creates moderate organizational and regulatory friction that prevents complete substitution, though it does not legally mandate human execution of the data-analysis step itself.
Adoption barriers	claude-sonnet-5	4/5	Civil engineering analysis often requires a licensed PE to review and stamp findings for safety and regulatory compliance, creating strong professional liability and licensing barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI-powered analysis of large datasets of maps, photographs, and surveys costs a fraction of manual review by licensed engineers. The per-unit cost of automated feature extraction and classification is orders of magnitude cheaper than human labor, though some oversight costs apply.
Cost vs. human wage	claude-sonnet-5	3/5	AI-assisted image/data processing tools reduce time on data extraction, but licensing, specialized software, and required human verification keep total costs roughly comparable to skilled engineer/technician labor rather than an order of magnitude cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature products exist today (GIS software with AI plugins, automated document analysis tools, aerial image processing systems used in production by engineering firms) that reliably extract and organize topographical and geologic data at scale. Minor gaps remain in handling ambiguous or handwritten legacy documents.
Technical feasibility today	claude-sonnet-5	3/5	GIS and computer vision products (e.g., automated feature extraction, remote sensing analytics) are deployed in practice, but reliability for engineering-critical topographic/geologic interpretation remains narrow and error-prone without expert review.

Prepare or present public reports on topics such as bid proposals, deeds, environmental impact statements, or property and right-of-way descriptions.

CI 34–51 · exposure 45 · augmentation 75 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Civil engineering firms are traditionally slower to digitize than IT/finance sectors; many still rely on templates and manual editing. While pilot use of AI writing tools is growing, production deployment at scale for official reports remains limited, constrained by liability concerns and conservative organizational practices.
Sector adoption velocity	claude-sonnet-5	2/5	Engineering and construction sectors have historically been slower to adopt AI tools compared to finance or software, though some firms are piloting AI-assisted documentation tools.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI is transformative for first-draft generation of bid proposals, environmental statement boilerplate, and property descriptions, significantly reducing write-time. Engineers remain in the loop to verify technical accuracy and legal compliance, but productivity gains on composition are substantial.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully speed up drafting, summarizing technical data, and formatting these reports, giving engineers a strong productivity boost while they retain responsibility for accuracy and sign-off.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can draft significant portions of these reports (bid proposals, property descriptions, standard environmental sections) with ~40–60% time savings, but human review and customization of legal/technical specifics, site-specific conditions, and regulatory compliance remain essential. The task requires domain knowledge integration across multiple data sources that AI handles incompletely without oversight.
Task automatability	claude-sonnet-5	3/5	AI can draft substantial portions of these reports (summarizing data, generating boilerplate text, formatting) but synthesis of engineering judgment, site-specific analysis, and final presentation to stakeholders still requires human involvement, so it doesn't fully meet the 50% end-to-end bar.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Professional liability and regulatory requirements mean engineers must sign off on and take responsibility for report accuracy, preventing full automation. However, AI-assisted drafting requires no licensing of the AI itself, only engineering review, so adoption friction is moderate rather than absolute.
Adoption barriers	claude-sonnet-5	4/5	These reports often require a licensed Professional Engineer's stamp/signature and must meet regulatory and legal standards (e.g., environmental law, property law), creating strong barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI-assisted drafting (few-shot prompting, template-filling) costs pennies to dollars per report section, while a civil engineer's hourly rate is $60–150+. Even accounting for oversight and revision, the cost ratio favors AI by 5–10×, especially for routine bid and property descriptions.
Cost vs. human wage	claude-sonnet-5	2/5	While AI drafting reduces some writing time cheaply, the need for licensed engineer review, data verification, and legal accuracy checks keeps overall costs closer to human-driven processes.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	AI writing and document-generation tools (GPT-based systems, legal document templates with ML) exist in production and can produce usable first drafts of reports. However, error rates on technical details, regulatory compliance, and site-specific accuracy remain material; human engineers must substantively review and edit before publication.
Technical feasibility today	claude-sonnet-5	2/5	AI writing/drafting tools are used ad hoc for report generation, but no deployed product reliably produces compliant bid proposals, deeds, or environmental impact statements without heavy engineer review and legal vetting.

Test soils or materials to determine the adequacy and strength of foundations, concrete, asphalt, or steel.

CI 25–25 · exposure 25 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Civil engineering and construction testing remains relatively traditional in adoption; while digital tools for data management are spreading, autonomous or semi-autonomous material testing and certification is still in pilot phases in most organizations.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction materials testing remain a physically-oriented, moderately digitized sector with slow AI adoption for core testing functions, though data analysis tools are gradually being piloted.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist engineers by automating data reduction, flagging anomalies in test results, and generating preliminary reports, improving review speed; however, the human engineer must remain central to validation and certification.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with analyzing test data, predicting material properties, flagging anomalies, and generating reports, improving efficiency while humans still perform and oversee physical testing.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can analyze laboratory test data and help interpret results, the physical testing itself—soil sampling, material preparation, equipment operation, and on-site assessment—requires manual intervention and professional judgment that current systems cannot perform end-to-end with 50% time savings at equal quality.
Task automatability	claude-sonnet-5	2/5	Physical soil/materials testing requires sampling, lab equipment, and hands-on procedures (e.g., Proctor tests, slump tests, compression tests) that AI cannot perform; only data analysis and interpretation portions are automatable.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Professional licensing (PE) and regulatory standards (ASTM, building codes) typically require a licensed engineer to certify test results and sign reports; liability and legal acceptance of AI-derived conclusions are significant friction.
Adoption barriers	claude-sonnet-5	4/5	Structural and geotechnical adequacy determinations often require licensed PE sign-off and adherence to building codes and safety regulations, creating strong liability and licensing barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The equipment, calibration, and regulatory compliance costs for material testing are substantial, and AI assistance on analysis alone does not reduce the labor cost of the technician or engineer performing the physical testing sufficiently to achieve cost parity.
Cost vs. human wage	claude-sonnet-5	2/5	Physical testing requires equipment, technicians, and site visits that AI cannot replace; cost savings are limited to interpretation and reporting steps, keeping overall costs comparable to human-led processes.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools exist for data analysis and report generation from test results, but no deployed product reliably performs the full cycle of material testing, evaluation, and professional certification autonomously; human engineers remain essential for test planning, execution, and approval.
Technical feasibility today	claude-sonnet-5	2/5	Some automated lab instruments and AI-assisted data analysis tools exist, but no deployed product performs the physical testing and adequacy determination end-to-end in production.

Direct or participate in surveying to lay out installations or establish reference points, grades, or elevations to guide construction.

CI 25–25 · exposure 25 · augmentation 63 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Civil engineering remains a moderately digitized field with heavy dependence on licensed professionals and on-site human judgment. While drone surveying is growing, autonomous layout direction remains rare in production, and regulatory/liability constraints slow broader adoption.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction are traditionally slow-adopting sectors for full automation of field tasks, though GPS/drone-assisted surveying tools are gradually being integrated.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist by processing survey data, automating calculation of grades and reference points, and generating layout visualizations that engineers then verify and deploy. This raises productivity in data processing phases while the engineer retains final judgment.
Augmentation potential	claude-sonnet-5	4/5	AI-enhanced surveying tools (drones, LiDAR, automated total stations, BIM integration) significantly speed up data collection and processing, letting engineers focus on directing and verifying layout work rather than manual measurement.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can process surveying data and generate plans, the physical task of establishing reference points and elevations on-site requires human judgment, equipment operation, and real-time environmental adaptation. AI cannot independently perform the full workflow at 50% time savings without substantial human oversight.
Task automatability	claude-sonnet-5	2/5	Field surveying and directing crews to establish physical reference points, grades, and elevations require on-site presence, physical measurement, and real-time judgment that current AI cannot perform end-to-end; software can assist calculations but not the physical survey or direction of personnel.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Surveying and establishing grades legally requires licensed professional engineers or surveyors in most jurisdictions; liability for grade/elevation errors in construction is high and often contractually assigned to the licensed human. This creates hard adoption barriers.
Adoption barriers	claude-sonnet-5	4/5	Surveying for construction often requires licensed surveyors or engineers to certify grades and reference points, with legal and liability implications tied to construction accuracy and safety, creating strong regulatory and professional barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Surveying equipment, field personnel, and licensed professionals are relatively low-cost compared to the infrastructure AI would need to reliably and safely execute field layout. Integration and liability costs push AI expenses closer to human labor costs.
Cost vs. human wage	claude-sonnet-5	2/5	Survey-grade equipment and software have high capital and calibration costs, and human oversight/direction is still required, so cost savings versus a licensed surveyor/engineer are modest at best.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No production system reliably performs end-to-end surveying layout autonomously. AI-assisted surveying tools exist (processing drone data, automating calculations), but deployed solutions still require licensed surveyors for final verification and field placement decisions.
Technical feasibility today	claude-sonnet-5	2/5	Robotic total stations, GPS/GNSS survey equipment, and drone-based photogrammetry are deployed but still require human operators, setup, and interpretation, especially for directing installation layout in construction contexts.

Identify environmental risks and develop risk management strategies for civil engineering projects.

CI 25–25 · exposure 25 · augmentation 63 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Adoption of AI for environmental risk assessment remains limited and primarily in pilot/exploratory phases. The engineering and construction sectors are slower to digitize core technical judgment tasks, and regulatory conservatism slows replacement of human-led environmental review processes.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction sectors show relatively slow AI adoption compared to information/finance industries, with most current use limited to pilot tools for document review or hazard mapping.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist engineers by automating literature review, screening regulatory databases, organizing historical project data, and flagging known hazard patterns, thereby accelerating the scoping phase. However, the core strategic judgment and site-specific synthesis still requires the human engineer's leadership.
Augmentation potential	claude-sonnet-5	4/5	AI tools can meaningfully assist by scanning regulations, prior incident data, and environmental datasets, helping engineers identify risks faster and draft mitigation frameworks, while engineers retain judgment and sign-off.
Task automatability	claude-haiku-4-5-20251001	2/5	Identifying environmental risks requires domain expertise, site-specific knowledge, and nuanced judgment across regulatory, geological, and ecological factors. While AI can assist in document review and pattern matching, developing comprehensive, context-sensitive risk management strategies still requires significant human expertise and on-site assessment.
Task automatability	claude-sonnet-5	2/5	AI can surface relevant environmental regulations, hazards, and precedents but synthesizing site-specific risk and designing mitigation strategies requires professional judgment, site knowledge, and accountability that current systems cannot fully replace.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Environmental risk assessment and management strategies often require sign-off by licensed professional engineers and must satisfy regulatory requirements (EPA, state environmental agencies, local permits). Liability exposure and legal accountability for missed risks create strong organizational and legal barriers to full automation.
Adoption barriers	claude-sonnet-5	4/5	Environmental risk assessments for civil projects often require licensed PE sign-off and are subject to regulatory review (e.g., environmental impact statements), creating strong liability and licensing barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Current AI systems require substantial human oversight, validation, and supplementation by licensed environmental engineers. The all-in cost (tool subscriptions, human review, integration) typically approaches or exceeds the cost of traditional preliminary engineering assessment by qualified professionals.
Cost vs. human wage	claude-sonnet-5	2/5	AI can cheaply generate draft risk lists or literature reviews, but the bulk of cost is in engineer time for validation, site assessment, and liability-bearing sign-off, keeping overall cost comparable to human-driven work.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some AI tools exist for preliminary environmental screening and hazard identification (e.g., document analysis, literature review), but no deployed product reliably performs end-to-end environmental risk identification and strategy development for civil projects at production scale. Existing systems have significant gaps in site-specific assessment and regulatory integration.
Technical feasibility today	claude-sonnet-5	2/5	Some environmental compliance and risk-screening tools exist, but no deployed product independently identifies environmental risks and designs management strategies for civil projects at production reliability.

Conduct studies of traffic patterns or environmental conditions to identify engineering problems and assess potential project impact.

CI 25–25 · exposure 25 · augmentation 63 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Civil engineering and transportation sectors show moderate digitization but slower adoption of agentic AI; pilot projects for traffic monitoring exist, but production deployment of autonomous impact assessment is still limited, and organizational culture favors licensed engineer accountability.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and infrastructure planning are traditionally slow-adopting sectors with limited digitization of physical-world data collection and analysis workflows.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist by automating data processing, visualizing patterns, and flagging anomalies that engineers then review and interpret; however, the human engineer must remain in the loop for final problem diagnosis and impact judgment.
Augmentation potential	claude-sonnet-5	4/5	AI-powered traffic simulation, GIS analysis, and environmental data tools significantly speed up data crunching and scenario modeling, meaningfully boosting engineer productivity while judgment and sign-off remain human.
Task automatability	claude-haiku-4-5-20251001	2/5	Traffic pattern and environmental data collection can be partially automated via sensors and image processing, but the interpretive work of identifying engineering problems and assessing project impact requires domain expertise, contextual judgment, and integration of multiple data streams that current AI systems struggle to do end-to-end without human oversight.
Task automatability	claude-sonnet-5	2/5	AI can assist with data analysis and pattern detection in traffic/environmental datasets, but the full study requires field observation, stakeholder input, professional judgment, and site-specific synthesis that current systems cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Civil engineering projects require licensed professional engineers to conduct studies and sign off on findings due to public safety and liability implications; regulatory requirements and the need for professional seal/authorization create substantial barriers to full automation.
Adoption barriers	claude-sonnet-5	4/5	Traffic and environmental impact assessments often require PE stamps, regulatory compliance, and legal accountability for public safety and environmental outcomes, creating strong professional and regulatory barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for data collection and preliminary analysis exist but do not replace the core analytical work; human civil engineers are still required for the judgment-heavy portions, making all-in cost comparable to or higher than human labor alone.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools can reduce data processing time but the overall study still requires substantial licensed engineer time for interpretation, field verification, and liability-bearing conclusions, keeping costs comparable to human-led work.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Deployed products can automate data gathering and flag anomalies, but assessment of engineering problems and impact evaluation remain heavily manual; no mature product reliably performs the full task of problem identification and impact assessment without significant human review.
Technical feasibility today	claude-sonnet-5	2/5	Traffic modeling and environmental impact software exist and incorporate AI-assisted analytics, but deployed products don't autonomously conduct full studies or reliably assess project impact without heavy engineer oversight.

Design energy-efficient or environmentally sound civil structures.

CI 25–25 · exposure 25 · augmentation 75 · importance 3.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Design offices use energy simulation tools and generative design plug-ins, but adoption remains in the pilot-and-assist category; most projects still rely on traditional human-led design with AI as an optional modeling aid, not as a replacement for the licensed engineer role.
Sector adoption velocity	claude-sonnet-5	2/5	Engineering and construction sectors are historically slower AI adopters compared to information/finance industries, with pilots for design-assist tools more common than widespread production use.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-powered energy simulation, parametric design exploration, and code-compliance checking substantially augment engineers' productivity during design iterations, enabling rapid exploration of alternatives and evidence-based optimization while the engineer retains judgment and accountability.
Augmentation potential	claude-sonnet-5	4/5	AI-driven simulation, energy modeling, and generative design tools meaningfully speed up exploration of sustainable design options while engineers retain final judgment and sign-off.
Task automatability	claude-haiku-4-5-20251001	2/5	Current AI can assist with parametric design optimization and energy modeling simulations, but cannot independently conceive structural systems, navigate competing constraints (safety, cost, aesthetics, regulatory), or validate designs against building codes—tasks requiring specialized engineering judgment and legal accountability.
Task automatability	claude-sonnet-5	2/5	This requires integrated judgment across structural, environmental, regulatory, and site-specific constraints that current AI cannot fully own end-to-end; AI can assist with analysis and drafting but not autonomously deliver a certified sustainable design.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Civil structures must meet building codes, safety standards, and professional engineering stamps (PE licensure required in most jurisdictions); liability for structural failure creates strong legal barriers, and clients typically require licensed engineer sign-off, preventing full automation.
Adoption barriers	claude-sonnet-5	4/5	Civil engineering designs typically require a licensed PE to review and stamp drawings, and liability for structural/environmental failures creates strong regulatory and professional barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Energy modeling software and AI tools reduce analysis time but add licensing costs; the value lies in accelerating iterations within a design led by humans, not in replacing the engineer's fee for conceptual and regulatory work, making overall cost comparable or higher than a fully human-directed process.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools reduce some analysis time but licensed engineers must still validate, iterate, and stamp designs, so overall cost savings versus a full human-led design process are modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools exist for energy simulation and parametric optimization (e.g., Ladybug, plugins in CAD), but they operate within human-defined design spaces and require expert oversight; no deployed product performs end-to-end structural design of novel environmentally sound buildings reliably without substantial human direction.
Technical feasibility today	claude-sonnet-5	2/5	Some generative design and energy-modeling tools exist (e.g., sustainability analysis plugins), but no deployed product independently designs full environmentally sound civil structures reliably in production.

Plan and design transportation or hydraulic systems or structures, using computer-assisted design or drawing tools.

CI 20–25 · exposure 25 · augmentation 75 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While engineering firms actively use CAD and AI-assisted drafting tools, adoption of autonomous design (beyond assistance) remains slow due to professional licensing requirements, liability concerns, and the need for human sign-off on all public-facing infrastructure.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction sectors are historically slow adopters of AI compared to software or finance, with CAD/BIM tools gradually incorporating AI features but production-scale autonomous design still rare.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly augments civil engineers through intelligent CAD assistance, parametric design suggestions, code checking, and rapid iteration of alternatives, allowing engineers to explore more design scenarios and focus on judgment-intensive optimization and safety review.
Augmentation potential	claude-sonnet-5	4/5	AI-enhanced CAD and simulation tools meaningfully speed up iteration, clash detection, and preliminary design generation, significantly boosting engineer productivity while the engineer retains final design authority.
Task automatability	claude-haiku-4-5-20251001	2/5	While CAD tools can generate routine structural elements and hydraulic schematics, transportation/hydraulic system design requires significant human judgment on site constraints, safety codes, environmental impact, and trade-offs that current AI cannot handle end-to-end. AI can assist with drafting and preliminary layouts but cannot replace the iterative, multi-constraint optimization that engineers perform.
Task automatability	claude-sonnet-5	2/5	AI can assist with drafting, generating design alternatives, and running calculations, but the core engineering judgment, code compliance, and integration of site-specific constraints still require substantial human expertise, so full end-to-end automation with equal quality is not yet achievable.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Transportation and hydraulic system design is typically stamped and licensed by professional engineers who bear legal liability for safety and performance. Regulatory requirements (ASCE, NTSB, EPA standards) and the non-delegable responsibility for public safety create hard barriers to full automation.
Adoption barriers	claude-sonnet-5	5/5	Civil engineering designs for public infrastructure typically require a licensed Professional Engineer's stamp and legal accountability, creating a hard regulatory barrier to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI-assisted CAD tools reduce drafting time but integration, oversight, and required human engineering review keep total cost comparable to or higher than the human labor they displace, especially for complex systems requiring regulatory compliance.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools can reduce drafting time but licensed engineers must still validate designs, review calculations, and stamp drawings, so overall cost savings versus a qualified engineer's involvement are limited.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Deployed AI systems can support CAD tasks (line detection, parametric modeling assistance) but no production system reliably designs complete transportation or hydraulic systems autonomously. Most applications are narrow (e.g., component generation) and require substantial human validation and rework.
Technical feasibility today	claude-sonnet-5	2/5	CAD tools have AI-assisted features (generative design, parametric modeling) but no deployed product autonomously plans and designs full transportation or hydraulic systems reliably in production without engineer oversight.

Provide technical advice to industrial or managerial personnel regarding design, construction, program modifications, or structural repairs.

CI 20–25 · exposure 25 · augmentation 75 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Civil engineering and construction remain relatively low-adoption sectors for autonomous AI; most firms use AI for analysis support rather than replacing advisory roles. Professional liability concerns and the fragmented, project-specific nature of engineering work slow widespread adoption of AI-driven advice systems.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction sectors are traditionally slow adopters of AI compared to finance or software, with pilots more common than production deployment.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can significantly assist civil engineers by rapidly generating design alternatives, running structural simulations, automating calculations, and summarizing complex technical data. When used as a tool to augment engineering judgment rather than replace it, AI measurably speeds up advisory work and improves decision quality.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully assist engineers by summarizing codes, generating draft reports, running preliminary calculations, and flagging design issues, boosting productivity while the engineer retains responsibility.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can generate design suggestions and structural analysis based on inputs, providing contextual technical advice requires deep understanding of site-specific constraints, stakeholder needs, and judgment calls that involve significant human oversight. Current AI systems struggle with the nuanced consultation aspect and cannot reliably handle the full advisory workflow without substantial expert validation.
Task automatability	claude-sonnet-5	2/5	This requires synthesizing site-specific structural knowledge, judgment, and liability-bearing advice that current AI cannot reliably generate end-to-end without heavy human verification.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Professional liability, legal sign-off requirements, and engineering licensing laws create hard barriers: in most jurisdictions, design and structural advice must ultimately come from or be stamped by a licensed professional engineer. Regulatory frameworks directly restrict substitution of unlicensed AI for professional judgment.
Adoption barriers	claude-sonnet-5	5/5	Structural and safety-related engineering advice typically requires a licensed Professional Engineer to sign off, creating a hard legal barrier to AI substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI infrastructure (GPT, specialized engineering models, integration) costs remain comparable to or exceed the cost of a junior engineer's time for advisory work, especially when accounting for oversight and liability risk mitigation. The advice quality and liability exposure make cheap automation unlikely.
Cost vs. human wage	claude-sonnet-5	2/5	While AI queries are cheap, the oversight and liability review needed to validate engineering advice keeps effective cost comparable to or only modestly below a licensed engineer's time.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools exist for structural analysis and design suggestions (e.g., FEA software, design assistants), but no deployed product reliably handles the full advisory consultation task—which requires interpreting vague requirements, negotiating tradeoffs, and communicating to non-technical managers. Existing systems perform narrowly scoped subtasks, not end-to-end advice.
Technical feasibility today	claude-sonnet-5	2/5	AI tools can assist with drafting reports or summarizing codes, but no deployed product independently provides authoritative technical advice on structural repairs or design modifications in production.

Inspect project sites to monitor progress and ensure conformance to design specifications and safety or sanitation standards.

CI 16–25 · exposure 17 · augmentation 75 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Construction and civil engineering remain relatively slow adopters of full automation; most AI adoption to date is pilots and assistive tools (drone data collection) rather than autonomous conformance assessment. Sectors are heavily human-in-the-loop and risk-averse, limiting production deployment velocity.
Sector adoption velocity	claude-sonnet-5	2/5	Construction remains a sector with relatively low digitization and slow AI adoption compared to information or finance industries, though drone-based monitoring is growing in pilot programs.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-powered drone imaging, automated progress tracking dashboards, and defect flagging systems substantially assist engineers in monitoring large sites and prioritizing areas for detailed inspection. These tools can compress time spent on data gathering and initial triage while engineers focus on judgment-intensive conformance validation.
Augmentation potential	claude-sonnet-5	4/5	AI-powered drones, photogrammetry, and computer vision tools increasingly help engineers track progress and flag deviations faster, meaningfully boosting productivity while the engineer retains final judgment.
Task automatability	claude-haiku-4-5-20251001	2/5	Visual inspection and progress monitoring can be partially automated using computer vision on drone imagery and photos, but assessing conformance to complex design specifications and identifying subtle safety/sanitation violations require contextual judgment and decision-making that current AI struggles with reliably. At best, AI can flag potential issues for human review, not meet the 50% time-saving threshold independently.
Task automatability	claude-sonnet-5	1/5	This requires physical presence at a construction site to visually inspect work, conditions, and compliance, which current AI systems cannot perform independently.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Civil engineers are responsible for safety and design compliance; legal liability and professional licensing requirements mean an engineer must typically certify conformance or at minimum sign off on findings. Regulatory standards (building codes, OSHA) reinforce human accountability, creating strong adoption friction.
Adoption barriers	claude-sonnet-5	4/5	Civil engineering inspections often require a licensed professional engineer to certify conformance and safety compliance, creating legal liability and licensure barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	While drone flights and image processing costs are falling, the integration overhead, site setup, and human expert review required to validate AI findings make all-in costs comparable to or potentially higher than a single site visit by an engineer. Savings accrue only at scale with repeated, standardized inspections.
Cost vs. human wage	claude-sonnet-5	2/5	Deploying drones, sensors, and AI analysis software adds equipment and integration costs, and still requires a qualified engineer to interpret findings and make judgment calls, so savings are modest relative to the human wage baseline.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Drone-based visual data collection and object detection exist in production, but end-to-end reliable conformance assessment across safety and sanitation standards remains limited to narrow, well-defined scenarios. Most deployments are assistive rather than fully autonomous, with significant error rates on complex or novel site conditions.
Technical feasibility today	claude-sonnet-5	2/5	Some drone/camera-based progress monitoring and computer vision tools exist for construction sites, but they supplement rather than replace physical inspection by a licensed engineer.

Design or engineer systems to efficiently dispose of chemical, biological, or other toxic wastes.

CI 13–20 · exposure 20 · augmentation 50 · importance 2.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Adoption of AI for waste system design is minimal; the sector remains heavily regulated, risk-averse, and dependent on individual PE credentials and liability. Small firms and public agencies dominate this domain with low digitization and heavy reliance on bespoke expert judgment, making automation adoption extremely slow.
Sector adoption velocity	claude-sonnet-5	2/5	Civil/environmental engineering is a moderately digitized but conservative, liability-heavy sector where AI adoption for specialized design tasks remains in early pilot stages.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI tools can usefully assist engineers by automating regulatory compliance checks, running contamination transport simulations, and summarizing technical literature, raising design efficiency. However, the core task—synthesizing site constraints, treatment options, and safety—remains human-led, limiting productivity gains to particular subtasks rather than transforming the overall process.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with literature review, preliminary calculations, regulatory research, and drafting design documents, improving engineer productivity on parts of the task while judgment and sign-off remain human.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with simulations, material selection, and regulatory compliance checking, designing waste disposal systems requires integrating site-specific geology, local regulations, chemical interactions, and novel engineering constraints that demand human expert judgment and accountability. Current AI cannot independently produce a complete, legally defensible design at >50% time savings compared to an engineer.
Task automatability	claude-sonnet-5	2/5	This requires site-specific engineering judgment, regulatory compliance, and integration of complex chemical/biological processes that current AI cannot autonomously design end-to-end; AI can assist with calculations and drafting but not full system design.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Significant legal and regulatory barriers exist: engineers must be licensed (PE), designs must be stamped by a responsible engineer, liability for failures rests on the signing professional, and regulatory agencies (EPA, state environmental boards) require documented professional accountability. A human engineer must legally supervise and approve any waste disposal design.
Adoption barriers	claude-sonnet-5	5/5	Waste disposal system design typically requires a licensed Professional Engineer's stamp and sign-off, with strict environmental and safety regulations creating hard legal barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI inference for auxiliary tasks (simulation, literature search) costs far less than human labor, but the engineer's role in design synthesis, risk assessment, and liability sign-off remains non-substitutable. The all-in cost of human engineering with AI assistance is still substantially cheaper than attempting to automate the role entirely given the need for expert oversight.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools can reduce time on calculations and documentation, but the overall design still requires substantial licensed engineer oversight, keeping costs comparable to human-led work with modest AI-driven savings.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed AI system reliably performs end-to-end toxic waste disposal system design in production. AI tools exist for narrower tasks (e.g., fluid dynamics simulation, regulatory database queries) but they require substantial expert interpretation and cannot replace the engineering synthesis required for safety-critical waste systems.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously designs toxic waste disposal systems in production; this remains a highly specialized engineering task requiring licensed professional judgment and site assessment.

Direct engineering activities, ensuring compliance with environmental, safety, or other governmental regulations.

CI 3–25 · exposure 13 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While engineering firms use compliance software and AI-assisted document review, actual autonomous direction of regulatory activities remains rare in production. Sector adoption of AI for this specific task is slow due to liability concerns and the critical human-in-the-loop requirement.
Sector adoption velocity	claude-sonnet-5	2/5	Civil engineering and construction sectors have historically slow AI adoption for core directive/compliance functions, though some digital tools are piloted for compliance checking.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist by scanning regulations, flagging potential non-compliance, suggesting applicable standards, and organizing documentation—meaningfully boosting a civil engineer's efficiency in compliance review and coordination activities. The human remains responsible for final decisions and sign-off.
Augmentation potential	claude-sonnet-5	3/5	AI can help by flagging regulatory changes, checking documents against codes, and summarizing compliance requirements, aiding but not replacing the directing engineer.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can help identify regulatory requirements and flag compliance gaps, the task inherently requires human judgment to interpret regulations, weigh tradeoffs, and make final decisions on engineering activities. Current AI cannot reliably direct the full scope of activities with legal accountability, limiting automatability well below the 50% time-savings threshold.
Task automatability	claude-sonnet-5	1/5	This is a supervisory/leadership task requiring judgment, accountability, and stamped professional sign-off; AI cannot direct people or bear regulatory responsibility today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Regulatory compliance, environmental sign-off, and safety authority typically require licensed professional engineers to certify and direct activities. Legal liability, professional licensing requirements, and mandatory human accountability create substantial barriers to full automation or delegation.
Adoption barriers	claude-sonnet-5	5/5	Professional engineering licensure, legal liability for regulatory compliance, and government mandates require a human engineer of record to direct and certify this work.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for regulatory compliance checking and document review are relatively affordable, but the human civil engineer's oversight and final judgment remain essential and expensive. The combined cost of AI plus mandatory expert review approaches or exceeds the cost of a human alone.
Cost vs. human wage	claude-sonnet-5	1/5	A licensed PE's directive and liability role cannot be substituted by AI inference costs; human oversight remains mandatory regardless of AI cost savings on subtasks.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Tools exist to assist with regulatory research and compliance checking (e.g., document analysis, gap identification), but no deployed product reliably performs end-to-end direction of engineering activities with regulatory oversight. Most production use is supportive rather than autonomous, and liability exposure prevents full deployment.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product directs engineering teams or assumes compliance oversight responsibility; at best AI tools support document review, not directing activities.

Develop or implement engineering solutions to clean up industrial accidents or other contaminated sites.

CI 3–25 · exposure 13 · augmentation 50 · importance 2.9/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Environmental and remediation engineering operates in heavily regulated sectors with strong preference for credentialed professionals; adoption of AI is limited to supporting tools (modeling, reporting) rather than autonomous solution development. Pilots exist but production automation is minimal.
Sector adoption velocity	claude-sonnet-5	2/5	Civil/environmental engineering and remediation sectors are physically grounded and slow to adopt AI for core technical decision-making, though software tools for modeling are used.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist with contamination modeling, site data analysis, regulatory research, and report generation, improving a civil engineer's productivity on technical and analytical phases while the engineer retains judgment on design decisions and regulatory strategy.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with contaminant transport modeling, literature review of remediation technologies, and report drafting, but core site-specific engineering design remains human-led.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with contamination modeling, risk assessment, and solution design, the task requires site-specific investigation, regulatory compliance decisions, and adaptive engineering judgment that current systems cannot perform end-to-end. Field work, equipment selection, and remediation oversight remain heavily human-dependent.
Task automatability	claude-sonnet-5	1/5	This requires physical site assessment, hazardous materials handling, on-site engineering judgment, and hands-on remediation design that cannot be executed end-to-end by AI systems today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers exist: environmental regulations often mandate licensed PE sign-off, liability for failed remediation is asymmetric (errors are costly and visible), and agency/client requirements for qualified human oversight are embedded in permitting and compliance frameworks.
Adoption barriers	claude-sonnet-5	5/5	Environmental remediation typically requires a licensed Professional Engineer's stamp, regulatory approval (EPA/state agencies), and legal liability for public health and safety outcomes.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for modeling and analysis reduce some preliminary design costs, but the overall task—spanning site characterization, regulatory consultation, equipment procurement, and adaptive field management—remains largely human labor-intensive and cheaper to execute with civil engineers than current AI solutions could provide.
Cost vs. human wage	claude-sonnet-5	1/5	AI cannot substitute for the licensed engineering work, site investigation, and liability-bearing sign-off involved, so there is no meaningful cost substitution today.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some narrow components (contamination simulation, site mapping from sensor data) have deployed products, but no integrated system reliably handles the full scope of solution development and implementation across diverse site conditions. Real-world remediation requires human expert judgment and regulatory sign-off.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product performs contamination remediation engineering; this remains fundamentally a human field-engineering and regulatory-compliance task.

Manage and direct the construction, operations, or maintenance activities at project site.

CI 3–21 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Construction remains a low-digitization sector with fragmented, small firms; while drone monitoring and scheduling tools are emerging, production adoption of autonomous site management is minimal and adoption velocity is slow.
Sector adoption velocity	claude-sonnet-5	2/5	Construction is a traditionally low-digitization, physical sector with slow AI adoption for site management functions, though software use for scheduling is growing.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist with schedule optimization, real-time monitoring dashboards, and document management, usefully reducing administrative overhead and improving visibility, but the human engineer remains essential for command decisions and safety judgment.
Augmentation potential	claude-sonnet-5	3/5	AI-based project management, scheduling, and monitoring tools (e.g., progress tracking via drones/sensors, BIM analytics) can meaningfully assist decision-making even though the engineer must direct operations in person.
Task automatability	claude-haiku-4-5-20251001	2/5	Site management involves real-time coordination of people, equipment, and safety conditions that require on-location judgment and dynamic decision-making. While AI can assist with scheduling and document review, the core task—presence-based oversight, conflict resolution, and adaptive resource allocation—remains fundamentally manual.
Task automatability	claude-sonnet-5	1/5	On-site management requires real-time physical presence, coordination with crews, equipment, and safety oversight that current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Construction site management has regulatory requirements (OSHA compliance, licensed engineer sign-off on critical decisions), liability exposure (safety incidents), and legal mandates for responsible human oversight, creating hard barriers to full automation.
Adoption barriers	claude-sonnet-5	5/5	Licensed Professional Engineer sign-off, safety liability, and legal responsibility for construction oversight require a human of record on site.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	AI tools that support site management (cameras, scheduling software) still require substantial human oversight and do not reduce the need for a salaried civil engineer on-site, making the all-in cost comparable to or higher than hiring the professional directly.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute performing this managerial/physical-directive role, so cost comparison favors the human engineer entirely.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed AI systems reliably manage live construction sites end-to-end. Scheduling software and monitoring cameras exist, but they address only narrow subtasks; a human civil engineer must still make final operational decisions and handle emergencies.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product manages or directs physical construction operations; existing tools are limited to scheduling, monitoring, or reporting support, not directing site activities.

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.