Millwrights

49-9044.00

Median wage $65,700/yr40,330 employed (US)Rank #855 of 923 scored · top 93% by substitution

Install, dismantle, or move machinery and heavy equipment according to layout plans, blueprints, or other drawings.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution12

Exposure7

Augmentation30

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

23 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%8

panel mean rating 1.3/5 → substitution pressure 8/100

Technical feasibility todayw 20%5

panel mean rating 1.2/5 → substitution pressure 5/100

Cost vs. human wagew 15%5

panel mean rating 1.2/5 → substitution pressure 5/100

Adoption barriersw 20%inverted — strong barriers lower the score34

panel mean rating 3.6/5 (barrier strength) → substitution pressure 34/100

Sector adoption velocityw 10%6

panel mean rating 1.2/5 → substitution pressure 6/100

Task breakdown (23 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.

Operate engine lathe to grind, file, and turn machine parts to dimensional specifications.

CI 10–82 · exposure 45 · augmentation 38 · 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	Manufacturing has undergone decades of CNC adoption; most industrial shops now rely heavily on automated lathes and grinding, though small job shops and specialized work retain manual operation.
Sector adoption velocity	claude-sonnet-5	1/5	Manufacturing and skilled trades sectors show slow AI adoption for physical hands-on tasks, with automation coming via CNC machining rather than AI systems, and millwright work remains highly manual and physical.
Augmentation potential	claude-haiku-4-5-20251001	3/5	CNC setup and tool-change assistance, digital measurement feedback, and AI-aided tolerance prediction can help millwrights work faster and more accurately, though the core cutting operation is increasingly autonomous.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with generating specifications, calculating dimensions, or troubleshooting via documentation, but offers minimal direct assistance to the hands-on physical operation of grinding, filing, and turning parts.
Task automatability	claude-haiku-4-5-20251001	5/5	CNC lathes and automated grinding/filing systems can perform dimensional turning, grinding, and filing end-to-end with precision tolerances, delivering >50% time savings compared to manual operation while meeting or exceeding quality specs on repetitive parts.
Task automatability	claude-sonnet-5	1/5	This requires physical manipulation of machinery, tactile feedback, and dexterity to operate a lathe on physical workpieces—current AI has no capability to perform physical manufacturing tasks.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Capital equipment investment and worker retraining create organizational friction, but no licensing barrier or legal requirement prevents automation of this task; unions and apprenticeship traditions add some resistance.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement specifically for lathe operation, but precision machining carries real error costs, quality control requirements, and typically requires trained skilled tradespeople with oversight.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated CNC machining costs a small fraction of skilled millwright labor per part, especially at volume, making it orders of magnitude cheaper when amortized across production runs.
Cost vs. human wage	claude-sonnet-5	1/5	AI systems cannot perform this physical task at all, so there is no viable cost comparison; any automation would require expensive CNC/robotic equipment, not AI software.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	CNC lathes and automated machining systems are mature, deployed at scale in manufacturing, though they require setup and still need human oversight for complex or custom specifications and tool changes.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product operates lathes or performs physical machining; this remains squarely in the robotics/automation domain, not general AI, and even robotic CNC automation differs fundamentally from AI performing manual lathe work.

Install robot and modify its program, using teach pendant.

CI 7–30 · exposure 13 · augmentation 50 · importance 3.8/5 · 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	Although manufacturing is digitizing, actual autonomous robot installation and programming by AI remains limited to pilots; millwrights and technicians remain the dominant mode of deployment in production environments across most sectors.
Sector adoption velocity	claude-sonnet-5	2/5	Industrial maintenance and robotics installation sectors show slow, uneven AI adoption compared to information-based professional services.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-assisted design tools, simulation software, and automated code generation can meaningfully accelerate program planning and reduce manual teach-pendant work, though a human expert must remain in the loop for installation verification and adaptive troubleshooting.
Augmentation potential	claude-sonnet-5	3/5	AI-assisted programming interfaces, simulation tools, and diagnostic software can help millwrights plan and verify robot programs, improving productivity on the programming portion of the task.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI could assist with aspects of program modification, installing physical robots and using a teach pendant requires real-time spatial reasoning, mechanical troubleshooting, and hands-on calibration that current AI cannot reliably execute end-to-end without human oversight and physical manipulation.
Task automatability	claude-sonnet-5	1/5	This is a physical installation task combined with hands-on programming via a teach pendant, requiring manual manipulation of hardware and physical positioning that AI cannot perform end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers exist: robot installation and modification often require licensed technicians or certification, liability for equipment damage is high, and regulatory safety compliance for industrial machinery demands human accountability and sign-off.
Adoption barriers	claude-sonnet-5	3/5	While not licensed in the same way as medical/legal work, safety protocols, physical access requirements, and specialized trade skills create meaningful friction against remote or automated substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI systems capable of assisting with robotic programming and simulation are specialized and costly, while millwright labor remains relatively inexpensive; the total cost of AI tooling and integration is not yet an order of magnitude cheaper than a trained technician.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing physical installation and pendant-based teaching, so cost comparison favors the human by default.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably handles full robot installation and teach-pendant programming autonomously; existing industrial robots require human technicians to execute these tasks, though some CAD software and simulation tools provide design support.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product installs industrial robots or handles teach pendant programming autonomously; this remains a manual skilled-trade task requiring physical presence.

Move machinery and equipment, using hoists, dollies, rollers, and trucks.

CI 5–30 · exposure 13 · augmentation 38 · importance 4.4/5 · 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	Manufacturing and heavy industry sectors adopt automation slowly for this task; most mills still rely on human operators and traditional hydraulic/mechanical equipment. Autonomous movement systems remain niche, with adoption concentrated in large integrated facilities rather than typical millwright work environments.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and millwright trades are a low-digitization, physical-labor sector with minimal AI/robotics adoption for this kind of task currently.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-assisted load calculation, path planning, and equipment diagnostics can improve efficiency and safety for human operators, and modern hoists increasingly include sensors and automation aids. However, augmentation is limited to advisory and control assistance rather than transformative productivity gain.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with planning logistics, load calculations, or route planning, but offers little direct enhancement to the physical act of moving machinery.
Task automatability	claude-haiku-4-5-20251001	2/5	While hoists and dollies can be partially automated, the task requires navigation of complex shop floors, precise positioning of heavy equipment, and real-time adjustment to obstacles—capabilities that current robotic systems lack reliably. Only narrow, pre-planned movements in controlled environments approach 50% time savings today.
Task automatability	claude-sonnet-5	1/5	Physically moving heavy machinery with hoists, dollies, and trucks requires manual dexterity, real-time spatial judgment, and physical strength that current AI systems cannot replicate end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	OSHA regulations and workplace safety standards mandate human oversight of machinery movement and load handling; liability for equipment damage and worker injury creates strong asymmetric error costs. Insurance and legal frameworks typically require a qualified human operator to sign off on critical lifts.
Adoption barriers	claude-sonnet-5	3/5	While not licensed work per se, safety regulations (OSHA rigging/hoisting rules) and liability for damaging expensive equipment create meaningful procedural barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Specialized heavy-lifting automation (cranes, hoists) is capital-intensive and requires significant integration. For general-purpose movement tasks, the all-in cost of autonomous systems, safety infrastructure, and oversight typically exceeds the loaded wage of a millwright.
Cost vs. human wage	claude-sonnet-5	1/5	Specialized robotics/automation for moving heavy irregular machinery would require expensive custom equipment far exceeding the cost of a millwright's labor for this task.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some industrial lifting equipment has basic automation, but end-to-end deployment of autonomous systems for general machinery movement remains research-stage or pilot-only. Production systems struggle with variability in equipment types, facility layouts, and safety-critical error costs.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously performs full-scale industrial equipment relocation; robotic material handling remains narrow and research/pilot stage for irregular heavy machinery.

Lay out mounting holes, using measuring instruments, and drill holes with power drill.

CI 5–30 · exposure 13 · augmentation 38 · importance 4.1/5 · 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	Millwright sectors (heavy manufacturing, utilities, on-site assembly) are relatively low-digitization, with small to medium firms dominating. Adoption of automation has been slow compared to white-collar sectors, with custom fabrication remaining labor-intensive.
Sector adoption velocity	claude-sonnet-5	1/5	Millwright work is a physical trade with low digitization and minimal AI/robotic adoption; industrial maintenance trades lag far behind information-sector automation trends.
Augmentation potential	claude-haiku-4-5-20251001	3/5	Digital calipers, laser layout tools, and mobile measurement apps can assist millwrights in marking hole locations faster and more accurately. However, the physical drilling and physical alignment tasks still require strong human control and adaptation to real-world conditions.
Augmentation potential	claude-sonnet-5	2/5	Digital layout tools, laser measuring devices, and CAD-based templates can assist planning and precision, but AI itself offers limited direct augmentation of the physical drilling action.
Task automatability	claude-haiku-4-5-20251001	2/5	Measuring and marking mounting holes could be partially automated with vision systems and coordinate data, but drilling itself requires physical robot arms with precise force control and real-time feedback. Current general-purpose AI systems lack reliable physical dexterity for the full end-to-end task at 50%+ time savings, though vision-based layout assistance exists in narrow industrial settings.
Task automatability	claude-sonnet-5	1/5	This is a physical layout and drilling task requiring precise manual manipulation of tools on machinery in varied real-world settings; no off-the-shelf AI or robotic system performs this end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Millwright work often requires on-site assembly and alignment that demands human judgment, physical presence, and accountability for fit and safety. Many manufacturing and heavy-equipment environments also have unionized labor protections and regulatory safety requirements that mandate skilled human oversight.
Adoption barriers	claude-sonnet-5	3/5	No licensing barrier specifically bars automation, but physical precision, safety requirements around power tools, and liability for structural/mechanical errors create real friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Industrial robotic arms capable of precision drilling are capital-intensive ($50k–$200k+) with integration costs, whereas a millwright charges loaded labor at $40–$60/hour. For one-off or small-batch hole drilling, AI remains more expensive than the human.
Cost vs. human wage	claude-sonnet-5	1/5	Achieving this via robotics would require expensive custom automation vastly more costly than a millwright's labor for this task, making AI/robotic substitution more expensive today.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Specialized industrial robots can drill holes with guidance, but they require extensive setup and custom tooling for each job configuration. No general-purpose, plug-and-play AI system reliably performs this task across varied equipment and geometries in production millwright environments today.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously lays out and drills mounting holes on industrial equipment in unstructured field environments; this remains outside current robotic capability at production scale.

Attach moving parts and subassemblies to basic assembly unit, using hand tools and power tools.

CI 5–30 · exposure 13 · augmentation 25 · importance 4.1/5 · 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	Millwrights work primarily in manufacturing and physical infrastructure—sectors with moderate, localized automation adoption. Most mills still rely on skilled human labor for assembly tasks; while some high-volume producers have invested in robotics, the overall sector is slow to displace this category of work.
Sector adoption velocity	claude-sonnet-5	1/5	Manufacturing/maintenance trades adopt automation slowly for physical mechanical assembly tasks, especially bespoke or heavy equipment work typical of millwrights.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI vision and guidance tools (e.g., augmented-reality overlays for part placement, torque specs) offer marginal assistance in planning, but the hands-on, real-time problem-solving nature of attachment work limits the scope of meaningful augmentation without removing the human from the core task.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with documentation, diagrams, or torque/spec lookups, but offers minimal direct assistance to the physical act of attaching parts and subassemblies.
Task automatability	claude-haiku-4-5-20251001	2/5	This task requires precise physical manipulation in three-dimensional space—aligning parts, fastening with hand/power tools, and verifying fit. While some component handling can be partially automated, the dexterity, positioning, and real-time adjustment needed for consistent quality across varied subassembly geometries remains beyond reliable current robotic systems without extensive custom setup.
Task automatability	claude-sonnet-5	1/5	This requires physical manipulation of mechanical parts with hand and power tools, dexterity, and fitting judgment that current AI systems cannot perform without embodied robotics far beyond generally available deployment.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety-critical assembly work in industrial settings often requires human sign-off, on-site troubleshooting, and immediate physical adjustment—functions that are hard to fully delegate to automation. Union and apprenticeship structures in the trades also create organizational and regulatory friction against full displacement.
Adoption barriers	claude-sonnet-5	3/5	No licensing strictly requires a human, but physical safety, equipment liability, and the need for hands-on adjustment during assembly create real practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Custom assembly robots are capital-intensive ($100k–$500k+), require integration and ongoing maintenance, and still need skilled technicians to reprogram for each product variant. For low-to-medium volume work typical of millwright roles, the all-in cost per assembly unit usually exceeds the loaded wage of a trained millwright.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI/robotic substitute at comparable capability, so any hypothetical automation solution (custom robotics) would be far costlier than a skilled millwright's wage for this task.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Specialized industrial robots exist for assembly in controlled, high-volume settings, but they require significant engineering per product line and struggle with variability in part geometry and alignment tolerances. No general-purpose deployed system reliably performs arbitrary millwright assembly tasks across typical plant environments.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs general millwright assembly attachment tasks in production; robotic assembly exists only for narrow, highly structured factory contexts, not this varied mechanical fitting work.

Conduct preventative maintenance and repair, and lubricate machines and equipment.

CI 5–26 · exposure 13 · augmentation 38 · importance 4.6/5 · 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	Manufacturing and maintenance sectors show moderate digitization but slow adoption of robotic maintenance automation. Predictive maintenance software has gained traction, but autonomous repair and lubrication systems remain niche pilots rather than widespread production deployments across the sector.
Sector adoption velocity	claude-sonnet-5	1/5	Manufacturing and industrial maintenance sectors have low AI adoption for physical tasks, relying on traditional mechanical/electrical technician labor with minimal automation of hands-on repair work.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-driven condition monitoring, anomaly detection from sensors, and maintenance scheduling tools meaningfully assist millwrights by flagging problems early and optimizing work order prioritization. These systems enhance human decision-making and reduce unplanned downtime, though the core hands-on work remains human-performed.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with predictive maintenance scheduling, diagnostics via sensor data analysis, and documentation, but offers little direct help with the physical lubrication and repair actions themselves.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can monitor sensor data and schedule maintenance, the physical acts of lubricating machines and conducting repairs require hands-on intervention, dexterity, and real-time environmental assessment that current AI systems cannot perform end-to-end. AI may assist in diagnostics and planning, but cannot achieve the ≥50% time-savings bar for the full task without human intervention.
Task automatability	claude-sonnet-5	1/5	This is a physical hands-on task requiring manipulation of heavy machinery, tools, and lubricants in real-world industrial settings, which current AI systems cannot perform.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety regulations, equipment-specific certifications, and liability for improper maintenance create legal and organizational friction. Many jurisdictions and equipment manufacturers require trained, licensed technicians to sign off on maintenance; there is high error-cost asymmetry if automated systems fail to detect or properly address mechanical issues.
Adoption barriers	claude-sonnet-5	3/5	While not always formally licensed, there are safety regulations, equipment liability, and physical dexterity requirements that create practical barriers to automation, though not strict professional licensure.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The hardware cost of robotic systems capable of performing complex maintenance tasks, combined with integration and safety oversight, exceeds the loaded wage of a skilled millwright performing the work manually. The capital and operational expenses make AI substitution uneconomical for this task today.
Cost vs. human wage	claude-sonnet-5	1/5	AI cannot perform the physical labor at all, so any comparison would require robotic hardware far exceeding the cost of a human millwright for this task today.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Predictive maintenance software exists and can forecast failures, but no deployed autonomous system reliably performs the actual lubrication, adjustment, or repair work itself. Mobile robots and manipulators remain experimental in unstructured industrial environments; production deployment of fully autonomous maintenance is not yet standard practice.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product physically conducts maintenance, repair, or lubrication of industrial equipment; this remains firmly in the domain of human technicians and robotics research, not general AI.

Troubleshoot equipment, electrical components, hydraulics, or other mechanical systems.

CI 5–26 · 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	Manufacturing and industrial maintenance sectors have slow-to-moderate AI adoption; while predictive maintenance pilots exist, autonomous troubleshooting remains nascent and adoption in production is rare, primarily confined to large enterprises.
Sector adoption velocity	claude-sonnet-5	1/5	Manufacturing maintenance and industrial trades are among the slowest sectors to adopt AI for physical hands-on tasks, with adoption largely limited to sensor-based monitoring rather than full troubleshooting automation.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist troubleshooting by suggesting diagnostic pathways, retrieving equipment manuals, or flagging anomaly patterns in sensor data, but the human millwright must still interpret findings, perform physical inspection, and make final judgment calls on repairs.
Augmentation potential	claude-sonnet-5	3/5	AI-powered diagnostic software, predictive maintenance sensors, and expert systems can help millwrights narrow down potential fault causes and access technical manuals faster, though the physical troubleshooting remains human-led.
Task automatability	claude-haiku-4-5-20251001	2/5	Troubleshooting complex mechanical and electrical systems requires physical inspection, contextual reasoning, and adaptive problem-solving that current AI struggles with. While AI can assist in diagnostic decision trees or manual lookup, end-to-end troubleshooting of diverse equipment types with 50% time savings at equal quality remains beyond today's systems.
Task automatability	claude-sonnet-5	1/5	Troubleshooting industrial machinery requires physical inspection, hands-on testing, and manipulation of hardware that current AI cannot perform end-to-end without a human physically present and executing diagnostics.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Liability and safety concerns are substantial: faulty troubleshooting diagnoses in industrial settings can cause injury, equipment damage, or production loss, creating strong disincentives to full automation. Additionally, licensed professionals (electricians, certified technicians) are often legally required to verify and sign off on critical repairs.
Adoption barriers	claude-sonnet-5	3/5	While not formally licensed in most jurisdictions, safety liability, physical access requirements, and specialized on-site tacit knowledge create substantial practical barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current AI diagnostic systems require significant integration, domain-specific training data, and human oversight to validate findings, making all-in costs comparable to or higher than a trained millwright's labor for equivalent troubleshooting outcomes.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-only substitute performing this physical diagnostic task, so any AI cost comparison is moot; humans remain the only cost-effective option.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably performs independent troubleshooting of industrial mechanical systems in production settings. AI diagnostic tools exist for narrow domains (e.g., specific equipment brands) but lack the generalization, physical sensing, and adaptive reasoning needed for real-world millwright work.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously diagnoses and resolves physical mechanical/hydraulic/electrical faults on industrial equipment; this remains research-stage robotics with no production deployment for millwright-scale troubleshooting.

Weld, repair, and fabricate equipment or machinery.

CI 5–25 · exposure 13 · augmentation 38 · 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	Millwright work is in manufacturing and maintenance sectors with slower digitization and is typically site-based, small-batch, and locally executed. Adoption of fully autonomous welding/fabrication remains limited to high-volume standardized settings; general repair and field fabrication adoption lags significantly.
Sector adoption velocity	claude-sonnet-5	1/5	Millwright work occurs in industrial/manufacturing maintenance settings with low digitization and minimal AI/robotics adoption for adaptive repair tasks.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-assisted tools (e.g., vision systems for weld planning, defect detection via imaging, or robotic arm guidance) can improve millwright productivity and safety on parts of the task. However, core judgment about material properties, fit, and real-time adaptation still relies heavily on human expertise, making augmentation moderate rather than transformative.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with diagnostics, documentation, or planning repairs, but offers little direct help with the hands-on welding and fabrication itself.
Task automatability	claude-haiku-4-5-20251001	2/5	Welding tasks require precise spatial reasoning, fine motor control, and real-time feedback adjustment in complex 3D environments. While robotic welding exists for highly repetitive, standardized jobs, the repair and fabrication aspect—which demands judgment about fit, alignment, and material properties—remains largely beyond current AI/automation without heavy specialization and operator oversight. Meaningful end-to-end automation at 50% time savings is not yet deployable for general millwright work.
Task automatability	claude-sonnet-5	1/5	Welding, repairing, and fabricating machinery requires physical manipulation, tactile feedback, and adaptive judgment in unstructured environments that current AI systems cannot perform end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Welding and fabrication of safety-critical equipment (pressure vessels, structural supports) often requires certified welders and may be governed by industry codes (ASME, AWS). Liability for failed welds, combined with licensing/certification requirements and the need for professional judgment on site, creates substantial legal and regulatory barriers to automation.
Adoption barriers	claude-sonnet-5	4/5	Welding certifications, safety regulations, and liability for structural/mechanical integrity impose strong requirements for qualified human oversight and sign-off.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Industrial welding equipment and robots, including infrastructure, integration, and the specialized programming required for non-standard jobs, remains expensive relative to skilled millwright labor for most repair and small-batch fabrication tasks. Only high-volume, standardized production achieves lower cost per unit.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this physical task, so any comparison favors the human millwright by default.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Welding robots exist in controlled factory settings for high-volume, standardized parts, but they do not reliably handle the diagnosis, repair, and adaptive fabrication that millwrights perform on heterogeneous equipment. No deployed system comprehensively handles site inspection, problem diagnosis, and custom fabrication without human expertise guiding each step.
Technical feasibility today	claude-sonnet-5	1/5	No deployed general-purpose product performs freeform industrial welding, repair, and fabrication; robotic welding exists only for fixed, pre-programmed factory lines, not adaptive repair work.

Bolt parts, such as side and deck plates, jaw plates, and journals, to basic assembly unit.

CI 5–23 · exposure 8 · augmentation 25 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Millwrighting is a physical, on-site trade in manufacturing and power generation—sectors with slower digital adoption. Custom bolt assembly automation is expensive and rare; most firms continue to rely on skilled workers.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial millwright work is a physical, low-digitization trade with minimal AI/robotics adoption for this specific type of variable heavy assembly task.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist marginally (e.g., by providing torque specs or bolt-location diagrams via computer vision), but the core task remains manual and operator-dependent, limiting meaningful productivity gains from current AI assistance.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with reference lookups, torque specs, or diagnostic guidance, but offers little direct help with the physical bolting and fitting process itself.
Task automatability	claude-haiku-4-5-20251001	2/5	Bolting physical parts requires precise positioning, torque control, and hand-eye coordination in 3D space. Current AI has no reliable way to physically manipulate hardware, and even with robotic arms, the variability of assembly units and bolt types makes full end-to-end automation without substantial setup unrealistic.
Task automatability	claude-sonnet-5	1/5	This is a physical assembly task requiring manual manipulation of heavy machine parts and hand tools; no current AI system can perform this bolting/fitting work end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	This task involves safety-critical structural assembly where bolting errors can cause equipment failure or injury. Liability, inspection requirements, and the need for a qualified worker to sign off on assembly quality create strong organizational and regulatory friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing barrier specifically for this task, but safety requirements, physical dexterity needs, and the custom/variable nature of heavy machinery assembly create practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Custom robotic systems capable of bolting are expensive to install, program, and maintain, often exceeding the loaded cost of a skilled millwright. Integration and oversight costs are substantial, making the total cost-per-task high relative to human labor.
Cost vs. human wage	claude-sonnet-5	1/5	Robotic systems capable of this dexterous, variable physical assembly with heavy irregular parts would cost far more than a millwright's wage, with no mature off-the-shelf solution available.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial product performs autonomous bolting of mill assembly units reliably in production. Specialized industrial robots exist but require extensive programming per configuration and are not off-the-shelf solutions for this class of task.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical bolting of heavy industrial components like jaw plates and journals; this remains firmly in the domain of human millwrights.

Dismantle machines, using hammers, wrenches, crowbars, and other hand tools.

CI 10–15 · exposure 0 · augmentation 25 · 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	1/5	Manufacturing and maintenance sectors show very limited robotic adoption for unstructured dismantling tasks; this work remains predominantly manual across the industry.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and millwright trades are a physical, low-digitization sector with minimal AI/robotics adoption for hands-on mechanical disassembly work.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with task planning or documentation, but hands-on tool use and physical problem-solving during dismantling offer minimal augmentation opportunity with current systems.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with documentation, diagnostics, or procedure lookup, but offers little direct assistance during the physical act of dismantling machines with hand tools.
Task automatability	claude-haiku-4-5-20251001	1/5	Dismantling machines with hand tools requires physical manipulation in unstructured, spatially complex environments. Current robotics cannot reliably grasp, manipulate, and disassemble diverse machinery with the dexterity and adaptability this task demands.
Task automatability	claude-sonnet-5	1/5	This is a physical manual disassembly task requiring hand-eye coordination, force application, and adaptive manipulation of varied machine components, none of which current AI systems can perform.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While there are no strict licensing barriers, the physical, safety-critical nature of dismantling and the need for site-specific adaptation create moderate organizational friction to automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing strictly requires a human, but safety regulations, liability for improperly dismantled heavy machinery, and the physical/spatial complexity create substantial practical barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Custom robotic systems capable of this task would cost orders of magnitude more than human millwright labor, with massive setup and customization required per job.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI/robotic substitute performing this task, so the human millwright remains the only cost-effective option; deploying robotics for this would be far more expensive than labor.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial system can autonomously dismantle machines at scale with hand tools today. This remains firmly in the research domain for robotic manipulation.
Technical feasibility today	claude-sonnet-5	1/5	No deployed robotic or AI product reliably dismantles industrial machinery with hand tools in production settings; this remains far beyond current robotics capability for unstructured mechanical disassembly.

Align machines or equipment, using hoists, jacks, hand tools, squares, rules, micrometers, lasers, or plumb bobs.

CI 5–14 · exposure 8 · augmentation 25 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Millwright work remains highly manual and physical, concentrated in manufacturing, maintenance, and construction—sectors with slow digitization. Adoption of autonomous alignment robotics is negligible; most mills and plants still rely entirely on skilled human technicians for this work.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and millwright trades are a low-digitization, physical-labor sector with minimal AI/robotic adoption for this kind of manual precision alignment work.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI-powered measurement and visualization tools (laser alignment guides, AR overlays, digital levels) provide some assistance in detecting misalignment, but they remain aids to human decision-making and physical adjustment rather than transformative productivity boosters for the core task.
Augmentation potential	claude-sonnet-5	2/5	Laser alignment tools and digital measurement devices already assist millwrights, but general AI systems add little beyond existing specialized instrumentation already in use.
Task automatability	claude-haiku-4-5-20251001	2/5	Aligning machines requires physical manipulation in 3D space with precision feedback, which current AI systems cannot perform autonomously. While vision systems can measure misalignment, the iterative adjustment with hand tools demands embodied robotics that remain unreliable at production speed and cost.
Task automatability	claude-sonnet-5	1/5	This is a physical precision alignment task requiring manual manipulation of hoists, jacks, and hand tools combined with tactile feedback and dexterity; no current AI system can perform this end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers exist: equipment alignment directly impacts safety and warranty; liability for misalignment falls on the party responsible for the work; most industrial sites require licensed or certified personnel to sign off on critical alignments; and customer preference for human expertise in high-consequence tasks.
Adoption barriers	claude-sonnet-5	4/5	Physical presence, specialized manual skill, safety considerations (heavy equipment, hoists, jacks), and often certification/apprenticeship requirements create strong barriers to any automated substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The capital cost of precision alignment robotics, integration, and ongoing maintenance far exceeds the loaded wage of a skilled millwright. Current robotic solutions cost tens of thousands to hundreds of thousands of dollars versus hourly labor of $30–50/hour for specialist trades.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-driven substitute performing this physical task, so any comparison would require expensive robotics far exceeding a millwright's wage for equivalent output.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial system performs end-to-end machine alignment autonomously today. Research robots exist for specific, controlled contexts, but production millwright alignment requires adaptability to diverse equipment, confined spaces, and real-time judgment that deployed systems do not reliably achieve.
Technical feasibility today	claude-sonnet-5	1/5	There are no deployed AI products performing physical machine alignment with hand tools and hoists; this remains firmly in the domain of skilled trades work with no robotic substitute in production.

Signal crane operator to lower basic assembly units to bedplate, and align unit to centerline.

CI 5–14 · exposure 8 · augmentation 25 · importance 4.6/5 · 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	Physical construction and manufacturing sectors historically show slow AI adoption; crane operations remain heavily regulated and dependent on licensed personnel, with minimal production-level displacement of signaling roles.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial millwright work in manufacturing/construction is a low-digitization, physical-labor sector with minimal AI or robotics adoption for this specific task.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could provide marginal assistance through real-time alignment visualization or laser-guided feedback systems, but the core task of signaling and decision-making remains deeply human-dependent and cannot be substantially enhanced without replacing the human role.
Augmentation potential	claude-sonnet-5	2/5	AI could marginally assist with digital alignment sensors, laser measurement tools, or predictive guidance, but the core signaling and physical alignment remains human-driven with limited AI augmentation currently in practice.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI could theoretically assist with visual alignment detection and signaling via autonomous systems, the task requires real-time coordination with a human crane operator and precise physical positioning on-site, which current autonomous systems cannot reliably perform end-to-end without human intervention.
Task automatability	claude-sonnet-5	1/5	This requires physical presence, real-time visual/tactile alignment judgment, and hand signaling during a physically hazardous crane operation—no current AI system can perform this physical coordination task.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: OSHA and industry safety regulations require a qualified human spotter/signaler for crane operations; liability and equipment damage risks create high error costs; and the task requires direct human presence and communication on the worksite.
Adoption barriers	claude-sonnet-5	4/5	Heavy machinery alignment carries significant safety and liability risk, often requires certified riggers/millwrights and adherence to OSHA crane safety protocols, creating strong barriers to non-human execution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current AI vision and signaling systems would require significant integration and continuous human oversight, making the all-in cost (hardware, software, integration, monitoring) higher than deploying a skilled millwright for this task.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute for this physical task, so any AI-based approach (e.g., robotic crane automation) would require costly specialized hardware far exceeding a millwright's wage for equivalent output today.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs crane signaling and alignment coordination autonomously in production environments; this remains a task requiring human judgment, spatial awareness, and direct coordination with equipment operators.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product exists that signals crane operators and physically aligns heavy machinery units to a centerline; this remains a manual, in-person skilled trade activity.

Fabricate and dismantle parts, equipment, and machines, using a cutting torch or other cutting equipment.

CI 5–14 · exposure 8 · augmentation 25 · importance 3.8/5 · 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	Millwright work is concentrated in small-to-mid-sized manufacturing and maintenance operations with lower digitization and capital budgets. Adoption of robotic torch-cutting remains limited to large, repetitive-task factories; most sectors where this task occurs lack the investment or process standardization to deploy automation.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and skilled trades are among the slowest sectors to adopt AI/robotics for physical fabrication tasks, with adoption dominated by traditional tools and manual labor.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could provide planning assistance (e.g., optimizing cut sequences or part layouts) or visualization aids, but the hands-on execution of torch cutting and dismantling leaves limited room for meaningful real-time AI assistance. The task remains tightly coupled to human sensorimotor control and judgment in the loop.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with planning, diagramming, or generating cut specifications, but offers minimal direct assistance during the physical execution of torch cutting and dismantling.
Task automatability	claude-haiku-4-5-20251001	2/5	Cutting torches and manual fabrication/dismantling require precise spatial judgment, real-time workpiece positioning, and reactive decision-making in physical space. While AI could assist with planning cuts, the actual torch operation—detecting part geometry, adjusting for irregularities, and safely executing cuts—remains heavily dependent on embodied human skill and judgment. End-to-end automation with ≥50% time savings is not achievable with current systems.
Task automatability	claude-sonnet-5	1/5	This is a physical hands-on task requiring manual dexterity with cutting torches and tools on heavy machinery; current AI systems cannot perform physical fabrication or dismantling work.
Adoption barriers	claude-haiku-4-5-20251001	4/5	OSHA regulations, workplace safety standards, and liability requirements create substantial barriers: automated torch cutting must meet strict safety certifications, and employers face liability if autonomous systems malfunction in high-hazard environments. Organizational friction and the expertise required to set up and maintain automated systems add friction.
Adoption barriers	claude-sonnet-5	4/5	Industrial safety regulations, welding/cutting certifications, and liability for structural and equipment integrity create strong barriers to any non-human execution of this task.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Robotic torch systems, when available, involve significant capital investment, maintenance, and integration costs that exceed the loaded wage of a skilled millwright for most small-to-mid-scale fabrication and dismantling work. The technology is expensive relative to manual labor in this context.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-driven substitute for this physical task, so any hypothetical automation (advanced robotics) would be far more capital-intensive than paying a millwright's wage.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI systems reliably perform autonomous torch cutting or equipment dismantling in production environments. Robotic systems exist for narrow, repetitive cutting tasks in controlled factory settings, but general-purpose fabrication and dismantling of varied equipment and machines is not a solved problem in commercial deployment.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical torch-cutting, fabrication, or dismantling of industrial equipment; this remains firmly in the domain of skilled trades work with no robotic substitutes in production.

Replace defective parts of machine, or adjust clearances and alignment of moving parts.

CI 5–10 · exposure 0 · augmentation 38 · importance 4.8/5 · 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	Manufacturing and facilities maintenance remain predominantly offline and labor-intensive; adoption of advanced robotics for millwright-level work is negligible in most sectors. Digitization is low, integration friction is high, and the highly distributed nature of millwright deployment (different sites, machines, problems) inhibits scale.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and millwright trades are physical, low-digitization sectors with minimal AI/robotics adoption for hands-on repair tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could marginally assist with diagnostic imagery analysis or documentation, but the core task—hands-on adjustment and replacement of parts—offers limited augmentation opportunity. Decision support tools for alignment calculations or spare-part lookup might help slightly, but gains are modest.
Augmentation potential	claude-sonnet-5	3/5	AI can assist via diagnostic tools, predictive maintenance analytics, and AR-guided repair instructions that help millwrights identify defects and alignment issues faster, though the physical execution remains manual.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical manipulation of machinery in diverse settings, precise spatial judgment about alignment and clearances, and diagnosis of mechanical problems that vary by machine type. Current AI systems lack embodied robotics at the dexterity level needed for such work; remote teleoperations remain experimental and do not achieve 50% time savings over skilled human millwrights.
Task automatability	claude-sonnet-5	1/5	This is a hands-on physical repair task requiring manipulation of heavy machinery parts, precise mechanical alignment, and tactile feedback that current AI systems cannot perform without a robotic embodiment capable of such dexterity.},
Adoption barriers	claude-haiku-4-5-20251001	4/5	Millwrights typically must hold trade certifications and licenses; many installations and repairs occur in unionized environments with strong labor agreements and safety regulations that mandate human oversight of critical machinery work. Liability for machine failure and worker safety creates strong barriers to full automation.
Adoption barriers	claude-sonnet-5	3/5	While not formally licensed in most jurisdictions, safety-critical machinery repair carries liability risk and often requires certified technicians per employer or industry standards, creating moderate friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current robotics platforms capable of precision mechanical work cost hundreds of thousands to millions and require extensive integration and human oversight. The loaded wage of a millwright is modest compared to the capital and operational cost of such systems, making AI significantly more expensive per task.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-driven substitute performing this physical task, so any comparison would require expensive specialized robotics far exceeding human labor costs.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs the full task of diagnosing defects, removing and replacing parts, and verifying alignment in production environments. Specialized robotic systems exist for narrow, repetitive subtasks in controlled settings, but not for the adaptive, diagnosis-heavy work millwrights do.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs autonomous replacement of machine parts or mechanical alignment adjustments in industrial settings; this remains far beyond current robotics capabilities for such varied, unstructured tasks.

Insert shims, adjust tension on nuts and bolts, or position parts, using hand tools and measuring instruments, to set specified clearances between moving and stationary parts.

CI 5–10 · exposure 0 · augmentation 25 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Millwright work is concentrated in manufacturing and heavy industry with low digitization, distributed across many small job sites, and strongly dependent on craft judgment; AI adoption remains minimal.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and millwright trades are a physical, low-digitization sector with minimal AI/robotics adoption for fine mechanical adjustment tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI tools could assist with documentation, clearance calculation, or work guidance, but the core task—hands-on mechanical adjustment with continuous feedback loops—offers limited augmentation potential without solving the underlying automation problem.
Augmentation potential	claude-sonnet-5	2/5	Digital measuring tools and diagnostic software can support precision measurement, but AI itself provides little direct assistance to the physical shimming and adjustment process.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires precise physical manipulation of mechanical parts in real-world space, insertion of shims, and fine adjustments guided by tactile feedback and visual inspection. Current AI cannot perform end-to-end physical assembly with the dexterity and adaptability this task demands.
Task automatability	claude-sonnet-5	1/5	This is a hands-on physical calibration task requiring fine motor manipulation of shims, fasteners, and measuring instruments in real industrial settings—far beyond current AI or robotic capability for general deployment.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Work often occurs on-site in existing machinery, requires real-time judgment about fit and clearance, and falls under occupational safety and equipment reliability standards that incentivize human accountability and verification.
Adoption barriers	claude-sonnet-5	3/5	No licensing mandate specifically requires a human for this micro-task, but safety-critical machinery alignment carries liability concerns and typically requires certified millwrights per employer/industry practice.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized robotics capable of precision mechanical adjustment cost significantly more than skilled millwright labor when amortized per task, and integration overhead is substantial; autonomous systems do not yet justify cost displacement.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI/robotic substitute performing this task at scale, so the human millwright remains the only cost-effective and available option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed autonomous system reliably performs shim insertion, bolt tensioning, and part positioning to tight mechanical tolerances in production environments. This requires embodied robotics with sophisticated manipulation and metrology integration, which remains research-stage for general millwright work.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs this precision mechanical alignment and adjustment task; it remains firmly in the domain of skilled human technicians with tactile feedback and judgment.

Assemble and install equipment, using hand tools and power tools.

CI 5–10 · exposure 0 · augmentation 25 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Millwrighting is a skilled trades occupation centered on physical, on-site work in industrial and manufacturing settings with limited digitization. Adoption of AI or automation in these sectors remains minimal and largely confined to large facilities with repetitive, controlled tasks.
Sector adoption velocity	claude-sonnet-5	1/5	Skilled trades and manufacturing installation work show minimal AI/robotics adoption for this kind of task; the sector is a laggard in autonomous physical task automation.
Augmentation potential	claude-haiku-4-5-20251001	2/5	While AI could theoretically assist with planning, CAD visualization, or troubleshooting guidance via remote expert systems, current deployments offer limited practical assistance for the hands-on assembly and installation work that defines the core of this task. Augmentation potential exists but is not yet realized at scale.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with diagnostics, manuals, or planning via AR/digital twins, but offers limited direct assistance to the hands-on physical assembly and installation process itself.
Task automatability	claude-haiku-4-5-20251001	1/5	Physical assembly and installation of equipment in real-world millwright contexts requires dexterity, spatial reasoning, and adaptability to site conditions that current AI systems cannot execute. While robots exist in controlled factory settings, the general-purpose assembly and on-site installation task described here remains beyond reach for automated systems today.
Task automatability	claude-sonnet-5	1/5	This requires physical manipulation of heavy machinery components in varied, unstructured environments—no current AI system can perform this physical assembly work end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Millwrights typically work in union environments with apprenticeship requirements and licensing standards; many sites require licensed, on-site personnel for equipment installation to ensure safety and liability compliance. Regulatory and contractual barriers to substitution are substantial.
Adoption barriers	claude-sonnet-5	3/5	No licensing strictly requires a human, but liability for improperly installed heavy machinery, safety regulations, and the need for on-site physical dexterity create substantial practical barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The capital cost of robotic systems capable of site-based assembly and installation, combined with integration and remote oversight, far exceeds the loaded wage of a skilled millwright. Current technology offers no cost advantage for this task.
Cost vs. human wage	claude-sonnet-5	1/5	Physical robotic systems capable of this flexible, precise heavy equipment installation would be far more expensive than a skilled millwright's wage, given costs of hardware, integration, and site-specific setup.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably performs general equipment assembly and installation with hand and power tools in the field. Specialized robotic systems exist only in narrow, structured manufacturing contexts, not in the diverse on-site millwright scenarios this task encompasses.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs full equipment assembly and installation with hand/power tools in industrial settings; robotics for such flexible, heavy-duty physical tasks remain research-stage or highly specialized to fixed factory lines.

Assemble machines, and bolt, weld, rivet, or otherwise fasten them to foundation or other structures, using hand tools and power tools.

CI 5–10 · exposure 0 · augmentation 25 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Millwright work occurs in construction, manufacturing, and maintenance settings—traditionally slow-adopting sectors with on-site physical constraints. Current adoption of AI or automation in this trade remains minimal, with work predominantly performed by human craftspeople.
Sector adoption velocity	claude-sonnet-5	1/5	Millwright work occurs in heavy manufacturing and industrial maintenance, sectors with low digitization and slow, physical-world automation adoption compared to information-based industries.
Augmentation potential	claude-haiku-4-5-20251001	2/5	While AI could potentially assist with planning assembly sequences or detecting defects via vision systems, the core task of hand assembly and fastening offers limited augmentation today. Current tools provide marginal productivity gains compared to the skill and autonomy required from the human operator.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with planning, diagrams, or torque/fastening specifications lookup, but offers minimal direct enhancement to the physical assembly and fastening process itself.
Task automatability	claude-haiku-4-5-20251001	1/5	Millwright assembly requires navigating complex 3D spatial layouts, selecting appropriate fastening methods for specific material combinations, and performing precision manual operations with hand and power tools. Current AI lacks the embodied dexterity, real-time sensorimotor feedback, and adaptive problem-solving needed to assemble machines and fasten components to foundations at scale.
Task automatability	claude-sonnet-5	1/5	This is physical hands-on assembly and fastening work requiring manipulation of heavy machinery components in variable industrial environments; no current AI system can perform this end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers exist including safety liability (assembly failures can cause equipment collapse or injury), union representation and apprenticeship requirements in many jurisdictions, need for human judgment on site-specific conditions, and customer preference for licensed/insured human millwrights who assume liability.
Adoption barriers	claude-sonnet-5	3/5	While not licensed like a trade such as electrician in all jurisdictions, safety regulations, certification for welding/rigging, and liability for structural failures create meaningful friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized industrial robots capable of assembly work are extremely expensive to purchase, program, and integrate compared to the hourly labor cost of skilled millwrights. Setup and customization costs per job would far exceed the cost of human labor for most applications.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute for this physical task, so the human remains the only cost-effective option; any robotic solution would require expensive custom engineering far exceeding a millwright's wage for one-off industrial installations.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI systems reliably perform end-to-end machine assembly or fastening operations in production environments. Robotic systems exist for narrow, highly controlled assembly tasks but lack the flexibility and adaptability required for the varied, site-specific nature of millwright work.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs full machine assembly, welding, riveting, and fastening to foundations autonomously in production settings; robotic welding exists only in highly structured, repetitive factory contexts, not general millwright work.

Shrink-fit bushings, sleeves, rings, liners, gears, and wheels to specified items, using portable gas heating equipment.

CI 5–10 · exposure 0 · augmentation 13 · 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	1/5	Millwright work occurs in manufacturing and industrial maintenance settings with low digital transformation penetration and strong reliance on skilled trades. No evidence suggests AI or automation adoption for shrink-fitting operations is underway in these sectors.
Sector adoption velocity	claude-sonnet-5	1/5	Millwright work occurs in heavy industrial and manufacturing maintenance settings, a sector with low AI/robotics adoption for physical manual tasks like this.
Augmentation potential	claude-haiku-4-5-20251001	1/5	AI offers minimal assistance for this task. While thermal imaging or measurement tools might support quality control in a limited way, the core work—heating, timing, positioning, and assembly—requires direct human control and cannot be meaningfully augmented by current AI systems.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with things like retrieving specifications, calculating thermal expansion tolerances, or documenting procedures, but offers minimal help with the actual physical heating and fitting process.
Task automatability	claude-haiku-4-5-20251001	1/5	Shrink-fitting is a highly specialized physical task requiring precise temperature control, tactile feedback, and real-time spatial judgment of fit and alignment. Current AI systems cannot operate portable gas heating equipment, sense thermal expansion in real time, or perform the manual dexterity required to position and secure components with the tolerances demanded in millwright work.
Task automatability	claude-sonnet-5	1/5	This is a physical hands-on task requiring manipulation of heavy machine parts and precise heating with portable gas equipment; no current AI system can perform the manual fitting and thermal application itself.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety regulations governing operation of portable gas equipment, combined with the need for a licensed millwright to certify fit tolerances and ensure structural integrity, create substantial legal and compliance barriers. Liability for equipment failure due to improper shrink-fitting is typically assigned to the qualified tradesperson.
Adoption barriers	claude-sonnet-5	3/5	While not licensed like a trade requiring formal certification for legal sign-off, this task requires specialized craft skill, safety training around gas heating equipment, and hands-on judgment that create practical barriers to any automated substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The capital cost of any robot capable of handling portable gas equipment and performing shrink-fitting would far exceed the loaded wage of a skilled millwright performing this task, with ongoing maintenance and setup costs adding significant overhead relative to manual labor.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI-based substitute for this physical task, so an all-in AI cost comparison is not applicable; the human millwright remains the only viable option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI or robotic system currently performs shrink-fitting with portable gas equipment in general millwright settings. While industrial robots exist for other tasks, the combination of mobile equipment operation, thermal sensing, and precise manual assembly under field conditions remains beyond production-ready automation.
Technical feasibility today	claude-sonnet-5	1/5	No deployed robotic or AI product performs shrink-fitting of bushings, sleeves, or gears in industrial settings; this remains a manual skilled-trade operation.

Construct foundation for machines, using hand tools and building materials such as wood, cement, and steel.

CI 5–10 · exposure 0 · augmentation 25 · 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	1/5	Foundation construction remains a largely traditional, on-site, labor-intensive craft with minimal AI/automation adoption; it is not digitized or amenable to remote or algorithmic optimization.
Sector adoption velocity	claude-sonnet-5	1/5	Skilled trades and industrial maintenance/construction are among the slowest sectors to adopt AI/robotics for physical fabrication work, with minimal production deployment.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with planning calculations and material specs, but current systems offer minimal real-time support for hands-on physical construction work in the field.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with planning, specifications, or generating construction drawings, but offers little direct assistance during the hands-on building and material work itself.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical site work involving precise excavation, measurement, and material handling in variable field conditions—capabilities not achievable by current AI systems without specialized robotics deployed on-site.
Task automatability	claude-sonnet-5	1/5	This is a physical construction task requiring manual dexterity, material handling, and on-site adaptation that current AI systems cannot perform; robotics for this specific unstructured task are not deployable today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: building codes and safety regulations require licensed, accountable human oversight of foundation work; liability falls on responsible parties; and unions typically control millwright roles.
Adoption barriers	claude-sonnet-5	3/5	No licensing law mandates a human specifically for this, but safety codes, structural liability, and site-specific judgment create real practical barriers to automation, though not strict legal ones.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The equipment, materials, and specialized robotics required to automate foundation construction would far exceed the loaded cost of a skilled millwright labor.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this physical task, so any AI-based approach (e.g., robotic construction) would be far more expensive than skilled human labor today.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI or robotics products reliably perform complete foundation construction end-to-end today; this remains in the physical domain requiring embodied agents.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously constructs machine foundations using hand tools and mixed materials; this remains firmly in the physical/manual domain unaddressed by current AI products.

Level bedplate and establish centerline, using straightedge, levels, and transit.

CI 0–10 · exposure 0 · augmentation 25 · 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	1/5	Millwright and heavy equipment installation is performed in dispersed, site-specific contexts with low digitization and high customization; adoption of AI automation in this sector remains negligible and is constrained by the physical and regulatory nature of the work.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial millwright and heavy equipment installation trades show minimal AI/robotic adoption for fine-grained physical alignment work, remaining a laggard sector for this kind of task.
Augmentation potential	claude-haiku-4-5-20251001	2/5	While AI-assisted measurement or digital modeling tools could marginally help a millwright plan or document alignment tasks, current systems offer limited real-time assistance during the hands-on leveling and centering process itself.
Augmentation potential	claude-sonnet-5	2/5	Laser levels, digital transits, and computer-assisted measurement tools can somewhat aid precision, but general AI systems offer little direct assistance to the core physical leveling task.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires precise physical manipulation in a real-world environment (leveling and aligning heavy industrial equipment) that current AI cannot perform autonomously. No off-the-shelf AI system can operate straightedges, levels, and transits or make real-time adjustments to a bedplate based on spatial feedback.
Task automatability	claude-sonnet-5	1/5	This is a hands-on physical calibration and alignment task requiring precise manual manipulation of heavy machinery components with instruments in a physical workspace; no current AI system can perform this end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Millwright work is typically performed by licensed tradespersons, and machinery installation often has regulatory and safety certifications that legally require qualified human sign-off and accountability for proper alignment and safety compliance.
Adoption barriers	claude-sonnet-5	3/5	No licensing law mandates a human specifically, but the physical nature of manipulating heavy equipment and the liability of misalignment create strong practical barriers to any automated substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current robotic systems capable of precision alignment work are prohibitively expensive to acquire and integrate compared to the hourly wage of a skilled millwright, making full automation economically unviable at present.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute performing this physical task, so the AI cost is effectively infinite relative to a millwright's wage.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product performs this physical, on-site millwright task today. The task demands embodied robotics with sophisticated spatial reasoning, feedback control, and adaptation to site conditions—capabilities not matured in production settings for this application.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical leveling and centerline establishment with straightedges, levels, and transits; this remains a purely manual skilled-trade task.

Dismantle machinery and equipment for shipment to installation site, performing installation and maintenance work as part of team.

CI 5–5 · exposure 0 · augmentation 25 · 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	1/5	Millwright work occurs primarily in manufacturing, construction, and maintenance sectors that lag in AI/automation adoption due to physical site constraints, customization requirements, and the need for skilled human judgment in unstructured environments.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and installation trades show minimal AI adoption; this is a physically-embodied, low-digitization sector.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with planning (disassembly sequencing, safety checks) or documentation, but current systems offer minimal productivity gain for the core physical and coordination demands of machinery dismantling. The task remains primarily human-dependent.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with planning logistics, generating checklists, or diagnostics documentation, but offers little help with the core physical dismantling and installation work.
Task automatability	claude-haiku-4-5-20251001	1/5	Dismantling heavy machinery requires physical manipulation, spatial reasoning, and real-time problem-solving in unstructured environments. Current AI cannot perform the embodied, dexterous work required at scale, and no general-purpose robotic systems reliably dismantle complex equipment end-to-end.
Task automatability	claude-sonnet-5	1/5	This requires physical disassembly, rigging, and hands-on manipulation of heavy industrial machinery, which is entirely outside current AI capability without embodied robotics.5-2050. No off-the-shelf system can perform this physical labor.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety regulations, liability for equipment damage, and insurance requirements create substantial barriers to automating machinery dismantling. Workers performing such tasks typically require trade certification, and human oversight of high-value equipment remains a legal and practical expectation.
Adoption barriers	claude-sonnet-5	4/5	Safety regulations, certification requirements for rigging/lifting, and liability for improper equipment handling create strong barriers to any automated substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Autonomous robotic systems capable of safe machinery dismantling are extremely expensive to acquire, maintain, and integrate, and would require significant site customization. The loaded cost far exceeds that of experienced human millwrights.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute for the physical labor involved, so any AI cost comparison is moot—human millwrights remain the only viable option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed products reliably perform machinery dismantling autonomously. Specialized industrial robots exist for narrow, pre-programmed tasks, but adaptable dismantling of varied equipment in field conditions remains beyond production AI systems.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical dismantling and installation of industrial machinery; this remains purely a human physical trade skill.

Connect power unit to machines or steam piping to equipment, and test unit to evaluate its mechanical operation.

CI 0–5 · exposure 0 · augmentation 25 · 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	1/5	Millwright work occurs primarily in manufacturing, utilities, and industrial sectors with low digitization of physical skilled tasks. These are capital-constrained, safety-regulated environments where adoption of automation for skilled manual work remains minimal and mostly limited to routine inspections.
Sector adoption velocity	claude-sonnet-5	1/5	Millwright work occurs in heavy industrial/manufacturing settings with low digitization and minimal AI/robotics adoption for physical installation tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist millwrights with diagnostic tools, equipment documentation lookup, or predictive maintenance alerts, but the core manual task of connecting and testing equipment offers limited augmentation opportunity since human judgment and physical dexterity remain central to safe execution.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with diagnostic data analysis or documentation during testing, but offers little help with the core physical connection and alignment work.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical manipulation of industrial equipment, precise mechanical connections, and hands-on testing in real-world environments—capabilities far beyond current AI and robotics. No current system can reliably perform the spatial reasoning, dexterity, and adaptive problem-solving needed to connect power units and steam piping.
Task automatability	claude-sonnet-5	1/5	This requires physical manipulation of heavy machinery, alignment of power units, connection of piping, and hands-on mechanical testing—none of which current AI systems can perform without embodied robotic capability far beyond present deployment.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong regulatory and safety barriers exist: equipment connections involving steam systems and power units are governed by industrial safety codes, OSHA standards, and equipment certifications that typically require a licensed millwright or qualified technician to perform and sign off on the work. Liability for equipment failure or safety hazards creates hard barriers to substitution.
Adoption barriers	claude-sonnet-5	4/5	Industrial equipment installation often involves safety regulations, certification requirements, and liability concerns around mechanical failures, creating strong barriers even if automation were technically possible.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of robotic systems capable of performing industrial equipment connections would vastly exceed the loaded hourly wage of a millwright, especially when accounting for integration, safety certification, and the need for custom programming per installation.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute for the physical labor involved, so any AI cost comparison is moot—human millwrights remain the only viable option and thus cheaper by default.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product performs end-to-end physical equipment connection and mechanical testing. While computer vision can inspect equipment, the core task of making power and steam connections with proper sealing and safety margins remains a skilled manual activity with no production automation.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical installation and mechanical testing of industrial power units; this remains purely a human physical trade skill.

Position steel beams to support bedplates of machines and equipment, using blueprints and schematic drawings to determine work procedures.

CI 0–5 · exposure 0 · augmentation 25 · 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	1/5	Millwrights work in physically constrained, project-based environments (manufacturing plants, construction sites) with low digitization; adoption of automation is minimal because the task demands real-time on-site physical work that robotics cannot yet perform economically.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial maintenance and millwright trades are a physical, low-digitization sector with minimal AI/robotic adoption for on-site heavy equipment installation tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist in blueprint interpretation or generate positioning checklists, but the core task of physically positioning beams remains human-dependent; augmentation value is limited to planning and documentation support rather than on-site task execution.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with interpreting blueprints or generating schematic overlays via image analysis tools, but this provides only marginal help to the core physical positioning task.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical manipulation of heavy steel beams in three-dimensional space, precise spatial reasoning from blueprints, and real-time environmental adaptation on job sites—capabilities far beyond current AI systems' reach without extensive custom hardware infrastructure.
Task automatability	claude-sonnet-5	1/5	This is a physical rigging and positioning task requiring manipulation of heavy steel beams on-site, which current AI cannot perform end-to-end; no software or robotic system does this autonomously today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Building codes, safety regulations, and liability law require a licensed or certified professional to sign off on structural load-bearing installations; equipment must meet OSHA and other safety standards, creating hard legal barriers to automation.
Adoption barriers	claude-sonnet-5	4/5	Heavy equipment rigging often involves safety regulations, certified rigger/millwright qualifications, and liability concerns for structural work, creating strong barriers to any automated substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Custom robotics capable of positioning structural steel beams with required precision and safety margins would cost orders of magnitude more than hiring a skilled millwright, including equipment, integration, and ongoing maintenance.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute for the physical labor and equipment operation involved, so AI cost is not comparable—human labor plus machinery remains the only viable option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed products can reliably perform the physical positioning of steel beams to support machinery bedplates; the task involves real-world robotics, load-bearing safety validation, and site-specific environmental constraints that remain in the research or prototype stage.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical beam positioning for machine bedplates; this remains firmly in the domain of human millwrights with cranes and hand tools.

Related occupations — Installation, Maintenance & Repair

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.