Electrical and Electronics Repairers, Powerhouse, Substation, and Relay
49-2095.00Median wage $103,020/yr20,720 employed (US)Rank #640 of 923 scored · top 69% by substitution
Inspect, test, repair, or maintain electrical equipment in generating stations, substations, and in-service relays.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution21
Exposure22
Augmentation48
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
15 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
0%
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%22
panel mean rating 1.9/5 → substitution pressure 22/100
Technical feasibility todayw 20%21
panel mean rating 1.8/5 → substitution pressure 21/100
Cost vs. human wagew 15%21
panel mean rating 1.8/5 → substitution pressure 21/100
Adoption barriersw 20%inverted — strong barriers lower the score24
panel mean rating 4.0/5 (barrier strength) → substitution pressure 24/100
Sector adoption velocityw 10%14
panel mean rating 1.6/5 → substitution pressure 14/100
Task breakdown (15 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.
Prepare and maintain records detailing tests, repairs, and maintenance.
66CI 60–71 · exposure 70 · augmentation 75 · importance 4.4/5 · click for rater detail
Prepare and maintain records detailing tests, repairs, and maintenance.
66| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Utility companies and electrical maintenance operations have digitized work orders and began adopting automated logging systems, but adoption varies by company size and legacy infrastructure. Production deployments of AI-assisted record management exist but are not yet industry-standard or deeply embedded. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Utility and power sector maintenance operations are historically slow adopters of AI tooling relative to information/finance sectors, with digitization of field records still maturing. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly assists human technicians by auto-populating fields, formatting records, suggesting relevant entries based on work type, and cross-referencing equipment history. This raises productivity and reduces clerical burden while technicians remain responsible for accuracy and compliance. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up drafting, summarizing, and structuring maintenance records from technician input, letting the human focus on verification and technical judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Record preparation and maintenance for equipment tests, repairs, and maintenance can be largely automated through structured data entry, OCR of handwritten notes, and AI-assisted documentation generation. However, some technical judgment in determining what to record and quality review may require human input, preventing a full 5 rating. |
| Task automatability | claude-sonnet-5 | 4/5 | Documenting tests, repairs, and maintenance is largely structured text/data entry that LLMs and voice-to-text/structured-form tools can draft or populate from technician notes with substantial time savings, though final verification still needed. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While record-keeping itself has no legal licensing requirement for the recording task, regulatory frameworks (FERC, NERC, utility standards) mandate that certain records be maintained and auditable, creating oversight requirements. Organizations often prefer human validation of technical accuracy, adding modest friction to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Utility maintenance records may have regulatory/compliance implications (NERC, safety audits) requiring accurate technician attestation, creating moderate barriers to full automation of authorship, though drafting assistance is unrestricted. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven documentation and record-keeping is substantially cheaper than human data entry and administrative time, with inference costs near zero and integration costs low for standard maintenance management systems. The loaded human wage for clerical/administrative record-keeping far exceeds the AI cost. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-assisted documentation (transcription, templated report generation) is far cheaper per record than a skilled technician's time spent writing narrative reports, though integration with legacy utility systems adds some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed systems today can reliably extract maintenance data from unstructured logs, populate standardized forms, and generate compliance documentation with minimal error. However, domain-specific technical accuracy and regulatory compliance validation often require human oversight in production environments. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Field service software with AI-assisted note generation and CMMS integration exists and is used in utilities, but adoption for this specific specialized equipment recordkeeping is uneven and often still manual or template-based. |
Notify facility personnel of equipment shutdowns.
53CI 43–64 · exposure 47 · augmentation 63 · importance 4.2/5 · click for rater detail
Notify facility personnel of equipment shutdowns.
53| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Utilities are adopting SCADA and alarm management systems, but full automation of notification logic remains rare in practice. Most facilities use semi-automated alerts with human gatekeeping; pilot and partial deployments are common, but production end-to-end automation lags. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Utility and power infrastructure sectors are traditionally slow adopters of new automation due to legacy systems, safety-critical culture, and regulatory oversight, despite the task itself being simple. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can pre-filter alarms, suggest recipient lists, draft notification messages, and log shutdown events automatically, significantly reducing the manual work required by facility operators. This augmentation is already deployed in some advanced control centers and measurably boosts operator productivity. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-driven alerting and automated dispatch tools can help ensure timely, consistent notifications and reduce missed communications, improving reliability for the technician managing the shutdown process. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Sending notifications is trivial to automate, but the task requires judgment about *when* to notify, *whom* to notify, and *how* to phrase alerts contextually—decisions that depend on equipment state, personnel roles, and facility-specific protocols. Current AI can generate messages but cannot reliably determine the correct notification trigger and recipient mapping in a live powerhouse environment. |
| Task automatability | claude-sonnet-5 | 3/5 | Sending shutdown notifications is a simple communication task that automated alerting systems (SCADA, email/SMS triggers) can already handle end-to-end for standardized events, though judgment calls on timing, phrasing, and escalation for complex situations still involve humans. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Safety-critical power infrastructure is regulated (NERC, FERC) and organizations are risk-averse about automating alerts that affect grid operations. Liability concerns and the need for human authorization on critical shutdowns create friction, though notification itself is not legally prohibited. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement to send a notification, but utilities have strict change-management and safety protocols that may require human confirmation before automated shutdown alerts are trusted as authoritative. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Once integrated, automated notifications cost pennies per alert versus the labor cost of a human monitoring and manually notifying staff. The upfront integration cost is moderate, but per-task cost becomes negligible at scale. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated notification via existing SCADA/monitoring infrastructure costs far less per event than having a technician manually notify staff, though some integration and monitoring overhead remains. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Notification systems (email, SMS, alert platforms) are mature, but integration with industrial control systems to automatically detect shutdowns and route alerts to the right personnel requires domain-specific configuration. Deployments exist but typically need significant customization and human oversight of alert rules. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Automated alerting and notification systems are mature and widely deployed in industrial and utility control systems today, reliably pushing outage/shutdown alerts to relevant personnel. |
Maintain inventories of spare parts for all equipment, requisitioning parts as necessary.
41CI 30–52 · exposure 42 · augmentation 63 · importance 3.5/5 · click for rater detail
Maintain inventories of spare parts for all equipment, requisitioning parts as necessary.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Utility and powerhouse operations are relatively slow adopters of cutting-edge automation; many rely on legacy inventory systems and conservative change management due to criticality and regulatory overhead, limiting real-world deployment of AI-driven requisitioning. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Utility and industrial maintenance sectors are generally slower adopters of advanced automation compared to information/professional services, though basic inventory software is common. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted demand forecasting, automated low-stock alerts, and supplier-matching tools can meaningfully assist technicians and inventory managers in planning and reducing manual data entry, though core judgment about critical spares remains human-driven. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Modern inventory management and procurement software significantly streamlines tracking, reordering, and alerts, meaningfully boosting efficiency while humans retain oversight of specialized part decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Inventory maintenance and requisitioning involve data entry, tracking, and some decision logic that current AI could partially automate, but the task requires ongoing physical inventory verification, supplier coordination, and judgment about necessary stock levels that exceed the 50% time-saving threshold for full automation today. |
| Task automatability | claude-sonnet-5 | 3/5 | Inventory tracking and requisition triggers can be automated via inventory management software and predictive analytics, but linking this to specialized electrical/electronic parts knowledge and physical stocktaking still needs human input. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Safety-critical infrastructure (powerhouses, substations) imposes regulatory requirements for equipment traceability, spare-parts accountability, and audit trails; physical verification and sign-off by licensed technicians are often mandated, creating legal and liability barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for inventory management itself, though purchasing authority and vendor relationships may require some organizational sign-off, creating mild friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Inventory management systems exist but require significant setup, integration with existing SCADA/maintenance systems, and ongoing human oversight; the all-in cost is often comparable to or exceeds the labor saved, especially in specialized powerhouse contexts. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Software-driven inventory systems reduce labor cost for tracking, but integration, procurement approvals, and physical counting still require paid staff time, keeping costs roughly comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Inventory management software and demand-forecasting tools exist in production, but they typically require human oversight of physical counts, supplier relationships, and exception handling in powerhouse/substation environments where equipment variety and criticality are high. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Enterprise inventory and procurement systems with automated reorder points are widely deployed and reliable, though they typically require human-configured thresholds and physical verification for specialized parts. |
Consult manuals, schematics, wiring diagrams, and engineering personnel to troubleshoot and solve equipment problems and to determine optimum equipment functioning.
25CI 25–25 · exposure 25 · augmentation 63 · importance 4.4/5 · click for rater detail
Consult manuals, schematics, wiring diagrams, and engineering personnel to troubleshoot and solve equipment problems and to determine optimum equipment functioning.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Electrical utilities and powerhouse operations are moderately digitized but conservative in safety-critical automation. Adoption of AI for unsupervised troubleshooting remains in pilot phases; deployment is limited by regulatory conservatism and skill-dependence. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Utility and power sector adoption of AI for field diagnostics is still nascent, with pilots more common than production deployment compared to faster-moving digital sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted schematic analysis, manual retrieval, and pattern matching on historical faults can meaningfully speed diagnosis, helping technicians narrow problem spaces and cross-reference solutions more quickly while they remain responsible for validation and repair execution. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by quickly searching schematics, manuals, and troubleshooting guides, and by suggesting likely fault causes, improving technician efficiency while they remain responsible for hands-on diagnosis and decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with reading and interpreting manuals and schematics, troubleshooting requires integration of visual inspection, hands-on testing, and real-time equipment diagnostics that current AI systems cannot fully automate. The consultation aspect and physical diagnostics create bottlenecks that prevent the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Portions like retrieving manual info or interpreting schematics can be AI-assisted, but hands-on diagnosis, physical inspection, and consultation with engineers on live high-voltage equipment cannot be automated end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Powerhouse and substation work is heavily regulated, licensed, and requires certified electricians to sign off on critical safety-related diagnostics. Safety liability and legal/regulatory requirements for human accountability on live electrical systems create substantial adoption barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Utility infrastructure work typically requires certified/licensed personnel, safety protocols, and regulatory compliance (e.g., NERC, OSHA), creating strong barriers to full automation or unsupervised AI action. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI document analysis and suggestion systems are inexpensive, but the human expertise required to validate diagnoses and perform physical testing remains the dominant cost. The overhead of oversight and potential re-work makes total cost comparable to or higher than human expertise alone. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply assist with document retrieval, but the core troubleshooting still requires skilled technician time and safety oversight, keeping overall cost comparable to or higher than pure human labor when integration and error risk are included. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs end-to-end troubleshooting of powerhouse and substation electrical equipment in production. AI tools can parse documents and suggest diagnoses, but real equipment diagnostics remain specialist work with unacceptable error rates if delegated to current systems alone. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some diagnostic assistants and manual-search tools exist, but no deployed product reliably troubleshoots powerhouse/substation relay equipment autonomously in production. |
Analyze test data to diagnose malfunctions, to determine performance characteristics of systems, or to evaluate effects of system modifications.
25CI 25–25 · exposure 25 · augmentation 63 · importance 4.2/5 · click for rater detail
Analyze test data to diagnose malfunctions, to determine performance characteristics of systems, or to evaluate effects of system modifications.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Utility and powerhouse sectors adopt digital tools slowly; AI adoption remains limited to data management and monitoring assists rather than autonomous diagnosis, reflecting both regulatory conservatism and the high cost of errors in mission-critical infrastructure. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Utility and power sector adoption of AI for hands-on technical diagnostics is slow due to legacy infrastructure, safety-critical culture, and conservative technology adoption cycles. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by organizing large test datasets, flagging anomalies, and suggesting diagnostic hypotheses, moderately raising technician productivity; however, human expertise remains essential for final diagnosis and decision-making, limiting transformative impact. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based analytics and pattern recognition can meaningfully assist technicians in interpreting test data, flagging anomalies, and comparing performance to historical baselines, improving diagnostic speed and accuracy. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can process and summarize test data, diagnosing malfunctions in complex electrical systems requires domain expertise, contextual judgment, and understanding of equipment-specific failure modes that current systems struggle with reliably. The task involves interpreting subtle patterns in specialized telemetry and making safety-critical decisions where errors carry high cost. |
| Task automatability | claude-sonnet-5 | 2/5 | Analyzing test data on electrical equipment involves specialized domain knowledge, physical context, and safety-critical judgment that current AI can assist with but not fully automate end-to-end.diagnostics require correlating physical sensor readings with equipment-specific behavior. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Powerhouse and substation work is heavily regulated (NERC, FERC, local utility codes), and diagnosis errors can cause catastrophic outages or safety hazards; liability and the need for licensed personnel to certify conclusions create hard legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Utility infrastructure work involves strict safety regulations, licensing, and liability concerns where a qualified technician must verify diagnoses before equipment modifications, creating strong sign-off barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for test data analysis have non-trivial integration and training costs, plus require expert human validation of outputs, meaning total all-in cost approaches or exceeds the loaded wage of a skilled electrical repairer without achieving full automation. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized diagnostic AI requires significant domain-specific integration, sensor infrastructure, and validation, making all-in costs comparable to or higher than skilled technician labor for now. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform end-to-end electrical system fault diagnosis at production scale; some tools assist with data visualization and anomaly flagging, but human experts must validate all conclusions. Existing AI solutions remain narrow and require significant human oversight in safety-critical contexts. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some diagnostic software and anomaly-detection tools exist for power systems, but they are narrow-scope decision-support tools rather than autonomous diagnosticians deployed at scale in substations. |
Run signal quality and connectivity tests for individual cables, and record results.
25CI 25–25 · exposure 25 · augmentation 50 · importance 3.8/5 · click for rater detail
Run signal quality and connectivity tests for individual cables, and record results.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | The utility sector adopts automation gradually; powerhouse and substation maintenance remain heavily staffed with skilled technicians. Few organizations report significant displacement of cable testing work, and adoption is primarily in data logging augmentation rather than autonomous test execution. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Utility and power infrastructure sectors are traditionally slow adopters of full automation for physical maintenance tasks, relying on established manual testing protocols and certified personnel. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted data logging, test scheduling, and anomaly flagging can meaningfully support technicians, and automated result recording reduces manual documentation burden. However, the physical and judgment-intensive aspects of test selection and cable setup remain human-driven. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Digital test equipment and software can automatically log results, flag anomalies, and generate reports, meaningfully speeding up the record-keeping portion of the task while the technician still performs the physical testing. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While cable testing equipment can be automated to run tests and log data, the task requires physical setup of test equipment, cable inspection, and contextual decision-making about which tests to run on which cables. Current AI/robotics cannot reliably handle the mechanical positioning and quality judgment aspects end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | The physical act of connecting test equipment to cables and running tests requires hands-on manipulation in a powerhouse/substation environment, though result recording and analysis could be automated; overall end-to-end automation is limited today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Electrical utility work in substations and power plants is heavily regulated; certified technicians must often perform or directly supervise cable testing for safety and liability reasons. Regulatory and safety standards (NERC, OSHA) create material barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Safety-critical electrical infrastructure typically requires qualified, often licensed technicians to perform and certify testing, with liability concerns around miswired connections affecting power reliability and safety. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automated cable testing systems require significant capital investment in robotics or specialized equipment integration, and the labor saved per task is modest. The loaded cost of setup and oversight often exceeds the wage savings for individual technicians in most operating environments. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Specialized test equipment and robotics for this environment are costly relative to a technician's time for what is often a routine, quick task; no clear order-of-magnitude cost advantage exists. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Specialized test equipment (cable testers, multimeters, signal analyzers) can be interfaced with data logging systems, but deployed automation for this task remains limited to narrow, pre-configured scenarios. Human technicians still predominate in real powerhouse and substation environments due to variability in cable types and testing protocols. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated cable testers and diagnostic tools exist and log data automatically, but full deployment for complex substation/relay cabling in production settings with no technician involvement is not demonstrated at scale. |
Schedule and supervise splicing or termination of cables in color-code order.
18CI 14–21 · exposure 20 · augmentation 50 · importance 3.9/5 · click for rater detail
Schedule and supervise splicing or termination of cables in color-code order.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Powerhouse and substation operations are capital-intensive, safety-critical infrastructure environments with strong union presence and regulatory constraints; AI adoption in this domain remains minimal despite digitization of some monitoring, with field technicians still predominant. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure maintenance sectors are slow to digitize and adopt AI for hands-on field supervision tasks, remaining a laggard sector with high physical and safety constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by automating color-code verification via image recognition, optimizing schedules, or flagging deviations from wiring diagrams, meaningfully improving the technician's efficiency on planning and QC steps while they retain hands-on and supervisory control. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with scheduling optimization, documentation, and tracking color-code compliance records, providing moderate productivity gains for the administrative portion of the task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Cable splicing and termination require precise physical manipulation and spatial reasoning in complex electrical systems; while AI could assist in scheduling and color-code verification via computer vision, actual splicing/termination remains a hands-on task AI cannot perform. Partial automation of scheduling alone falls well short of 50% time savings on the full end-to-end task. |
| Task automatability | claude-sonnet-5 | 2/5 | Scheduling could be partially automated, but supervising physical splicing/termination work in the field requires hands-on presence, judgment, and real-time quality verification that current AI cannot perform end-to-end.: |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Electrical powerhouse and substation work is heavily regulated under OSHA, NFPA, and utility-specific standards; a licensed electrician or qualified electrical worker must legally supervise and sign off on high-voltage splicing and termination, creating hard legal and liability barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | High-voltage electrical work is subject to strict safety regulations, certification requirements, and liability concerns, requiring qualified personnel to supervise and sign off on such work. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems capable of scheduling and supervising this work would require custom integration, on-site computer vision, and ongoing human oversight, making the all-in cost substantially higher than the loaded wage of a skilled technician who performs the work directly. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical supervision and safety-critical judgment involved, so the human remains necessary and AI offers no cost substitution for the core task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably automate the full task of scheduling and supervising cable splicing in the field; scheduling software exists but integrating it with real-time physical supervision and color-code verification at scale is not demonstrated in production systems. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product supervises or performs physical cable splicing/termination in substations; this remains a specialized skilled-trade activity requiring on-site human oversight. |
Test insulators and bushings of equipment by inducing voltage across insulation, testing current, and calculating insulation loss.
15CI 5–25 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail
Test insulators and bushings of equipment by inducing voltage across insulation, testing current, and calculating insulation loss.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Powerhouse and substation maintenance remains traditional, with limited digitization and slow adoption cycles. Utilities are conservative and heavily regulated; pilots of automated testing are rare, and production AI deployment in this safety-critical domain is minimal. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Electrical utility maintenance and repair is a physically-intensive, safety-critical sector with minimal AI/robotic adoption for hands-on high-voltage testing tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by automating insulation loss calculations from test data, suggesting anomalies, and streamlining data logging, helping a technician work more efficiently. However, the physical testing execution and safety judgment remain human-dependent, limiting augmentation upside. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with calculating insulation loss from collected data, flagging anomalies, and generating reports, but cannot perform the physical voltage induction and probing steps. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While inducing voltage and measuring current are automatable in principle, the task requires interpreting complex insulation loss calculations in context of equipment-specific safety protocols and judgment about equipment degradation. Current AI lacks the embodied capability to physically connect testing equipment and safely execute this safety-critical procedure end-to-end. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical manipulation of test equipment, hands-on connection to high-voltage apparatus, and interpretation within a physically hazardous environment—no AI system can perform the physical testing steps. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Powerhouse and substation work is heavily regulated under OSHA, NERC, and utility-specific licensing requirements; equipment testing typically requires a licensed electrician or engineer to execute and sign off on results. Liability for equipment failure or safety incidents creates hard barriers to autonomous operation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | High-voltage electrical work typically requires certified/licensed technicians, strict safety protocols (lockout/tagout, PPE), and utility regulatory compliance, creating strong barriers to any non-human performance. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference for calculation is cheap, but the task requires expensive specialized testing equipment setup, physical connection, and skilled human supervision for safety. The all-in cost of AI-assisted testing still requires the technician present, so cost savings are modest compared to loaded technician wages. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical test, so comparing costs is moot—human technicians with specialized test equipment are the only current option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools can perform insulation loss calculations and data analysis from logged measurements, but no deployed product reliably performs the full task of independent testing execution, voltage induction, current measurement, and diagnostic interpretation in production powerhouse environments without human technician oversight. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical insulation/bushing testing on powerhouse or substation equipment; this remains purely a manual, instrument-based field task. |
Schedule and supervise the construction and testing of special devices and the implementation of unique monitoring or control systems.
15CI 5–25 · exposure 13 · augmentation 50 · importance 3.9/5 · click for rater detail
Schedule and supervise the construction and testing of special devices and the implementation of unique monitoring or control systems.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Power utilities are risk-averse and operate under strict regulatory frameworks. While they adopt AI for predictive maintenance and asset management, they have been slow to automate supervisory roles on construction and testing of critical systems due to safety and regulatory concerns. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure sectors are slow-adopting, physically grounded, and highly regulated, showing minimal AI agent deployment in this domain. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist with work-order scheduling, test protocol generation, and documentation, improving a supervisor's ability to manage multiple projects and catch gaps in testing checklists. However, augmentation is limited to administrative and planning support, not judgment-critical aspects of live system supervision. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with scheduling optimization, documentation, simulation of control system logic, and monitoring data analysis, providing useful support even though the core supervisory and construction work remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with scheduling algorithms and generate testing protocols, the task requires on-site supervision, real-time judgment of construction quality, and live system testing that demand human presence and decision-making. Current AI cannot reliably supervise physical construction or validate unique monitoring systems end-to-end without extensive human oversight. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical construction, hands-on testing of specialized electrical devices, and on-site supervision of skilled trades work that AI cannot perform end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory oversight of electrical power systems is stringent (FERC, NERC, state utility commissions), and liability for failed systems falls on accountable human supervisors. Professional licensure (PE/electrical) is often required to sign off on critical infrastructure modifications, creating hard legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Power infrastructure work involves safety-critical systems, often requires licensed electricians/engineers and regulatory compliance (NERC, OSHA), and errors carry high liability, creating strong barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI can reduce scheduling and documentation costs, but the specialized expertise required to supervise unique monitoring and control system implementation commands high hourly rates, while AI solutions still require significant human oversight and validation, making the cost ratio unfavorable for full substitution. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical supervision and hands-on judgment involved, so there is no viable AI-only cost basis for comparison versus the skilled human technician/supervisor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably supervise construction and testing of specialized electrical devices and control systems autonomously. AI scheduling tools exist, but actual supervision—inspecting work, making go/no-go decisions on untested systems, managing contingencies—remains human-dependent in production environments. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product schedules and supervises physical construction and testing of custom electrical/monitoring systems; this remains firmly human-managed field work. |
Test oil in circuit breakers and transformers for dielectric strength, refilling oil periodically.
12CI 5–19 · exposure 13 · augmentation 38 · importance 3.9/5 · click for rater detail
Test oil in circuit breakers and transformers for dielectric strength, refilling oil periodically.
12| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Powerhouse and substation environments are traditional, capital-intensive, safety-focused operations with long equipment lifecycles and conservative upgrade cycles. Adoption of automated oil-testing and refilling systems remains minimal; utilities typically rely on trained technicians and scheduled maintenance protocols. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure maintenance is a low-digitization, physically-intensive sector with minimal AI/robotic adoption for hands-on equipment servicing. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted condition monitoring (predictive analytics on oil degradation trends, automated scheduling of tests, digital record-keeping) can help technicians prioritize maintenance and reduce manual record-keeping, but the core task of sample extraction and refilling remains human-supervised. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with predictive maintenance scheduling or analyzing test results/trends, but offers little direct help with the physical sampling and refilling process itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI-guided sampling and measurement of oil dielectric strength via instruments can be partially automated, the physical manipulation of circuit breakers and transformers—extracting samples, refilling oil, sealing components—requires on-site embodied robotics that is not yet reliable or cost-effective at scale. Current systems cannot independently perform the full end-to-end workflow with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical, hands-on task requiring sampling oil, using dielectric test equipment, and physically refilling equipment in the field; current AI systems cannot perform the physical manipulation involved. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Safety-critical infrastructure work on energized or de-energized high-voltage equipment is heavily regulated and typically requires licensed electricians or authorized technicians to perform or verify the work. Liability for equipment damage or failure, plus legal/regulatory sign-off requirements, create substantial barriers to unsupervised automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Utility infrastructure maintenance often requires certified technicians, safety protocols, and regulatory compliance (e.g., NERC, OSHA), plus physical access barriers that prevent easy AI/robotic substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Specialized robotic systems capable of safely handling high-voltage equipment and oils, plus integration and oversight, would be prohibitively expensive compared to the labor cost of a trained technician performing these tests and maintenance activities. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute for the physical labor involved, so the human technician remains the only viable option, making AI effectively infinitely more costly (no functioning alternative). |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Measurement devices and sensors for dielectric strength testing exist and are deployed, but the full task—including safe oil extraction, sample handling, refilling, and re-commissioning—lacks reliable automated solutions in production. Partial automation (data logging, flagging threshold breaches) is feasible; full substitution is not. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI/robotic product performs oil sampling, dielectric testing, and refilling of transformers/breakers autonomously; this remains a manual maintenance task performed by technicians. |
Repair, replace, and clean equipment and components such as circuit breakers, brushes, and commutators.
9CI 0–18 · exposure 8 · augmentation 25 · importance 4.0/5 · click for rater detail
Repair, replace, and clean equipment and components such as circuit breakers, brushes, and commutators.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | The electric utility sector, which comprises the primary employers of these repairers, has historically lagged in automation adoption due to safety requirements, regulatory frameworks, and the critical nature of infrastructure. Adoption of AI for this specific task in production remains minimal. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility maintenance and power infrastructure repair is a low-digitization, physically intensive sector with minimal AI/robotic adoption for hands-on repair tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist with diagnostics or documentation of equipment conditions, but the hands-on nature of repair, replacement, and cleaning work limits meaningful augmentation. Human expertise and physical presence remain central to task execution. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with diagnostics, predictive maintenance scheduling, or documentation lookup, but offers little direct help with the physical repair and cleaning actions themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While some parts of equipment inspection and cleaning could theoretically be assisted by robotic systems, the task requires fine motor control, spatial reasoning in complex electrical environments, and judgment about equipment conditions that current AI systems cannot reliably execute end-to-end. The high-risk nature of powerhouse environments and the need for safe handling of live or de-energized equipment makes autonomous performance infeasible. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical hands-on repair task requiring manipulation of heavy electrical equipment in industrial settings; no current AI system can perform physical repair, replacement, or cleaning of hardware. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Significant legal and regulatory barriers exist: electrical work on critical infrastructure typically requires licensed electricians, and liability for equipment failure is substantial. Safety regulations and the requirement for certified personnel to sign off on repairs create hard adoption barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Safety regulations, electrical licensing requirements, and high liability for utility infrastructure failures create strong barriers, though not always requiring a specific single sign-off like some professions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The specialized equipment, safety systems, and oversight required to automate this task would currently exceed the cost of trained electricians, especially considering integration, calibration, and the need for human verification in safety-critical environments. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-based substitute for the physical labor involved, so AI cost is effectively infinite relative to human labor for the manual component of this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products reliably perform the full scope of repair, replacement, and cleaning of circuit breakers, brushes, and commutators in powerhouse settings. While robotic arms exist in controlled manufacturing, they are not production-ready for the variability and safety requirements of this electrical infrastructure task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed products perform physical repair/cleaning of circuit breakers, brushes, or commutators; this remains firmly in the domain of skilled human technicians with robotics still research-stage for such tasks. |
Inspect and test equipment and circuits to identify malfunctions or defects, using wiring diagrams and testing devices such as ohmmeters, voltmeters, or ammeters.
7CI 0–14 · exposure 8 · augmentation 38 · importance 4.5/5 · click for rater detail
Inspect and test equipment and circuits to identify malfunctions or defects, using wiring diagrams and testing devices such as ohmmeters, voltmeters, or ammeters.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Powerhouse and substation work remains largely physical, field-based, and bound by strict safety and licensing regulations. Adoption of autonomous inspection AI in these sectors is minimal; pilots are rare and production deployment is virtually nonexistent. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure maintenance is a low-digitization, physically intensive sector with minimal AI/robotic deployment for hands-on equipment testing. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide limited assistance by auto-interpreting wiring diagrams or suggesting test sequences, but the physical inspection and real-time decision-making in hazardous environments require the technician to remain fully engaged, limiting meaningful productivity gains. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by analyzing sensor data, predicting failure points, or referencing wiring diagrams and manuals, improving diagnostic efficiency even though the physical testing remains human-performed. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could assist in analyzing wiring diagrams and interpreting test readings, the physical inspection of equipment and the hands-on use of testing devices requires embodied presence in the field. Current AI cannot autonomously perform the inspection component or safely operate in live electrical environments, limiting end-to-end automation to roughly 20-30% of the task. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical presence, hands-on use of test equipment, and interpretation of readings in context of physical equipment condition—no current AI system can perform the physical inspection or testing steps. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strict OSHA/NFPA regulatory requirements, licensing mandates for high-voltage work, liability asymmetry (failures can cause death/injury), and legal requirements that a licensed electrician must perform or sign off on safety-critical inspections create hard legal barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | High-voltage electrical work is heavily regulated with safety certifications, lockout-tagout procedures, and liability concerns that require qualified, often licensed personnel to perform and sign off on inspections. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The specialized equipment, safety compliance, integration overhead, and need for human oversight in hazardous electrical environments make AI-assisted inspection more expensive than traditional technician methods today. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical labor and equipment handling involved, so there is no viable AI-only cost comparison; a human technician remains necessary. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs in-person electrical equipment inspection and malfunction diagnosis in live powerhouse or substation environments. Computer vision for circuit board inspection exists in narrow lab settings, but real-world deployment in high-voltage, safety-critical contexts remains research-stage. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously inspects and tests powerhouse/substation equipment; existing tools are limited to diagnostic data analysis support, not physical execution. |
Construct, test, maintain, and repair substation relay and control systems.
7CI 0–14 · exposure 13 · augmentation 38 · importance 4.1/5 · click for rater detail
Construct, test, maintain, and repair substation relay and control systems.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Utilities are conservative, risk-averse organizations with long infrastructure lifecycles and strong regulatory constraints. Adoption of AI for field repairs and construction has been minimal; the sector remains dependent on skilled human technicians and incremental digital tools. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure maintenance is a low-digitization, physically-grounded sector with minimal AI agent deployment in the field for hands-on repair tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist with remote monitoring dashboards, predictive maintenance alerts, and diagnostic recommendations, but provides limited real-time augmentation during actual construction and hands-on repair work. Most value is in planning rather than active task execution. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with diagnostics, documentation, predictive maintenance analytics, and troubleshooting guidance, providing moderate productivity support while the physical work remains human-performed. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While some subcomponents like testing and diagnostics can be partially automated with specialized tools and software, the physical construction, on-site maintenance, and repair of substation systems require manual dexterity, spatial reasoning, and judgment in complex electrical environments. Current AI systems cannot handle the full end-to-end task at 50% time savings with equal quality. |
| Task automatability | claude-sonnet-5 | 1/5 | This is hands-on physical work involving construction, testing, and repair of high-voltage electrical equipment that requires physical manipulation, diagnostic judgment, and safety-critical decision-making that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Substation work is heavily regulated by utility commissions, NERC standards, and electrical codes that typically require licensed electricians to perform or certify work. Liability for grid failures, safety hazards in high-voltage environments, and mandatory human sign-off create hard legal barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Utility work on substations involves strict regulatory oversight (NERC, OSHA), licensing/certification requirements, high safety and liability stakes, and mandatory qualified-worker sign-off, making substitution essentially impossible without a licensed human. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The specialized hardware, sensors, and skilled technicians required for substation work remain significantly cheaper than developing and deploying autonomous systems capable of safe operation in high-voltage environments. Integration and safety oversight costs would exceed human technician wages. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so any AI cost comparison is moot; human labor with specialized training remains the only option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Diagnostic software exists for relay testing and monitoring, but no deployed AI system can autonomously construct, maintain, or repair physical substation hardware in real-world conditions. Products are limited to narrow aspects like anomaly detection in operational data, not the full task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product exists that autonomously constructs, tests, or repairs substation relay and control systems; this remains firmly in the domain of skilled human technicians. |
Open and close switches to isolate defective relays, performing adjustments or repairs.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail
Open and close switches to isolate defective relays, performing adjustments or repairs.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of automation in power-utility operations is slow for safety-critical physical tasks. Substations remain heavily staffed with human technicians, and regulatory/safety requirements strongly discourage substitution of autonomous systems for this work. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure maintenance is a physically demanding, highly regulated sector with minimal AI-driven automation of hands-on repair tasks to date. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide marginal assistance via diagnostic recommendations or decision support (e.g., identifying which relay is faulty), but the core task of physically opening/closing switches and performing repairs remains almost entirely human-driven. The augmentation potential is limited by the physical and safety-critical nature of the work. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with diagnostics, documentation, or troubleshooting guidance, but offers limited direct assistance for the physical switching and repair actions themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation in a high-voltage environment with safety-critical decisions about which switches to open/close. Current AI systems cannot perform physical actions on electrical equipment, and the decision logic depends on real-time inspection and safety protocols that demand human expertise and presence. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical manipulation of high-voltage switches and hands-on relay adjustment/repair in a hazardous substation environment, which current AI systems cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | High licensing and legal barriers exist: only qualified, licensed electricians are authorized to work on substation and relay equipment under electrical codes and safety regulations. Liability for errors is extreme (risk of electrocution, equipment damage, grid instability), creating hard regulatory and liability barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Utility work on energized/de-energized electrical equipment requires certified qualified personnel under strict safety regulations (e.g., NERC, OSHA lockout/tagout), making unsupervised automation legally and physically prohibited. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI has no meaningful role in the physical execution of this task, so comparison to human cost is not applicable. A technician must perform this work, and automation is not a cost-replacement option. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for the physical isolation and repair work, so AI cost comparison is effectively inapplicable and human labor remains the only option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can physically operate electrical switches or diagnose relay defects in situ. This remains firmly in the domain of human electrical technicians with specialized training and authority to work in substations. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product operates or repairs electrical switches and relays autonomously in production; this remains physical field work requiring human presence. |
Disconnect voltage regulators, bolts, and screws, and connect replacement regulators to high-voltage lines.
0CI 0–0 · exposure 0 · augmentation 25 · importance 3.7/5 · click for rater detail
Disconnect voltage regulators, bolts, and screws, and connect replacement regulators to high-voltage lines.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Utility companies remain highly regulated and risk-averse; physical infrastructure work on live or de-energized equipment has minimal automation adoption. The sector shows limited velocity toward autonomous or robotic replacement of licensed electrical work. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Utility and power infrastructure maintenance is a low-digitization, physical-labor sector with minimal AI/robotic adoption for hands-on high-voltage repair tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with pre-work diagnostics, voltage monitoring, or documentation, but these are peripheral to the core mechanical and electrical task. The hands-on physical work itself offers minimal augmentation surface for current AI. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with diagnostics, scheduling, or documentation around this task, but offers little direct augmentation for the physical act of disconnecting and connecting voltage regulators. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation of high-voltage equipment in live or de-energized substations, precise mechanical connections, and real-time safety assessment. Current AI systems cannot perform end-to-end physical disconnection, handling, and reconnection of heavy electrical components. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical, hands-on electrical task requiring manipulation of high-voltage equipment, bolts, and wiring in the field; current AI has no capability to perform physical disconnection/connection of hardware. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task has hard regulatory and legal barriers: only licensed electricians can work on high-voltage equipment, utilities face strict safety codes (OSHA, NERC) mandating human certification and sign-off, and liability for failures is severe. Automation of this task is legally restricted, not just organizationally discouraged. |
| Adoption barriers | claude-sonnet-5 | 5/5 | High-voltage electrical work is heavily regulated, requires licensed/certified electricians and safety protocols (lockout-tagout, arc-flash protection), and carries severe liability for errors, making non-human substitution essentially barred. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Specialized robotics capable of safe high-voltage work (if available) would cost orders of magnitude more than the loaded wage of a skilled electrical repairer, making this economically infeasible for routine maintenance tasks. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI/robotic substitute performing this physical task, so the human technician remains the only cost-effective option; specialized robotics would be far more expensive than a technician's wage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs high-voltage electrical equipment replacement in production environments. This requires specialized robotics with extreme precision, safety interlocks, and real-time electrical hazard detection—none of which exist as reliable commercial solutions in substation contexts. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed robotic or AI product performs live high-voltage regulator swaps; this remains far outside current automation, even research-stage robotics for this specific task are largely absent. |
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.