Stationary Engineers and Boiler Operators

51-8021.00
Median wage $78,620/yr28,250 employed (US)Rank #702 of 923 scored · top 76% by substitution

Operate or maintain stationary engines, boilers, or other mechanical equipment to provide utilities for buildings or industrial processes. Operate equipment such as steam engines, generators, motors, turbines, and steam boilers.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution19
Exposure19
Augmentation47

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

25 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

4%

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%18

panel mean rating 1.7/5 → substitution pressure 18/100

Technical feasibility todayw 20%20

panel mean rating 1.8/5 → substitution pressure 20/100

Cost vs. human wagew 15%19

panel mean rating 1.8/5 → substitution pressure 19/100

Adoption barriersw 20%inverted — strong barriers lower the score21

panel mean rating 4.1/5 (barrier strength) → substitution pressure 21/100

Sector adoption velocityw 10%13

panel mean rating 1.5/5 → substitution pressure 13/100

Task breakdown (25 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.

Weigh, measure, and record fuel used.

72

CI 65–79 · exposure 75 · augmentation 75 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5The industrial and energy sectors have rapidly adopted automated fuel monitoring and SCADA systems over the past 15–20 years. Large facilities (power plants, refineries, manufacturing) have mostly transitioned to continuous automated measurement; smaller facilities lag but adoption is accelerating.
Sector adoption velocityclaude-sonnet-52/5Industrial/utility sectors adopt automation but often on long capital cycles; many boiler plants still use manual or semi-manual logging especially in smaller facilities.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-driven fuel tracking systems substantially augment operator productivity by providing real-time visibility, alerts for anomalies, forecasting, and historical analytics. The operator remains responsible for interpretation and response, but their ability to monitor and optimize is transformed.
Augmentation potentialclaude-sonnet-54/5Automated meters and digital dashboards significantly reduce manual burden and improve accuracy of fuel tracking while operators retain oversight of the broader boiler process.
Task automatabilityclaude-haiku-4-5-202510014/5The task of weighing, measuring, and recording fuel is highly repetitive and procedural. Current AI systems with IoT sensors and automated logging can capture measurements directly and record them with minimal human intervention, achieving well over 50% time savings at equal or better accuracy than manual records.
Task automatabilityclaude-sonnet-54/5This is a straightforward data-capture and recording task that can be handled via automated sensors, meters, and data logging systems with minimal human involvement, meeting the time-saving threshold easily.
Adoption barriersclaude-haiku-4-5-202510013/5There is modest friction: regulatory requirements in some jurisdictions mandate certified operator sign-off on fuel records, and organizations may prefer human oversight for accountability. However, these are not hard legal barriers to automation itself, and many facilities have already moved to automated logging.
Adoption barriersclaude-sonnet-52/5No licensing requirement for measurement/logging itself, though some facilities may have compliance/record-keeping protocols requiring certified operator sign-off on logs.
Cost vs. human wageclaude-haiku-4-5-202510015/5Once sensor and logging infrastructure is in place, the marginal cost per measurement-recording cycle is negligible (fractions of a cent per transaction), vastly cheaper than the hourly labor cost of a stationary engineer manually weighing and recording fuel.
Cost vs. human wageclaude-sonnet-54/5Once sensors and logging infrastructure are installed, per-instance recording cost is negligible compared to a human manually weighing/measuring and logging fuel.
Technical feasibility todayclaude-haiku-4-5-202510014/5Automated fuel measurement systems (flow meters, weight sensors) integrated with SCADA and digital logging platforms are widely deployed in industrial facilities. These systems reliably perform the measurement and recording in production environments, though some manual verification or occasional calibration oversight remains common.
Technical feasibility todayclaude-sonnet-54/5Industrial fuel metering, flow sensors, and automated SCADA/data historian systems are mature and widely deployed in plants to log fuel consumption automatically.

Maintain daily logs of operation, maintenance, and safety activities, including test results, instrument readings, and details of equipment malfunctions and maintenance work.

43

CI 25–60 · exposure 45 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Boiler operations and stationary engineering remain traditional, physically-rooted sectors with aging workforces; adoption of AI-driven logging is slow and limited to large industrial facilities, with most operations still relying on manual or semi-automated systems.
Sector adoption velocityclaude-sonnet-52/5Industrial/facilities operations sectors are historically slow adopters of AI-driven automation relative to information/finance sectors, with digitization of maintenance logs still uneven across smaller facilities.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by auto-populating sensor data, flagging anomalies, and organizing logs, but the human operator must interpret readings, make judgment calls on equipment status, and ensure compliance—genuine productivity gain is moderate and localized to data entry and formatting stages.
Augmentation potentialclaude-sonnet-54/5AI-enabled sensors, automated data logging, and anomaly-flagging tools can substantially reduce manual burden and improve consistency of logs while the operator remains responsible for verification and judgment calls.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with log formatting and data entry from sensor readings, the task requires human judgment to interpret equipment status, identify anomalies, document safety concerns, and decide what constitutes a critical malfunction—judgments that cannot be fully automated end-to-end today without significant human oversight.
Task automatabilityclaude-sonnet-54/5Logging structured data like test results, instrument readings, and malfunction details is largely a data-entry/transcription task that AI-integrated systems (IoT sensors, SCADA integration, voice-to-text) can automate with high time savings once sensor data feeds are connected.
Adoption barriersclaude-haiku-4-5-202510014/5Regulatory frameworks (OSHA, boiler codes, EPA rules) typically require that a licensed operator sign and attest to safety logs; liability and safety-critical nature of documentation create hard barriers to full automation or substitution.
Adoption barriersclaude-sonnet-53/5Some regulatory frameworks require certified operators to verify and sign off on logs for safety and compliance (e.g., boiler inspection records), creating a documentation/liability barrier even if data capture is automated.
Cost vs. human wageclaude-haiku-4-5-202510012/5Automated logging solutions exist but require integration, calibration, and continuous human verification; the all-in cost of oversight and error correction often approaches or exceeds the cost of direct human logging, especially given liability sensitivity.
Cost vs. human wageclaude-sonnet-54/5Automated sensor logging and digital record-keeping systems are inexpensive to run compared to the operator's time spent manually transcribing readings, though initial sensor/integration setup adds cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some logging systems can auto-capture sensor data and generate templates, but no deployed product reliably documents maintenance work, interprets equipment malfunctions, and ensures safety-compliant record-keeping without human review and correction in operational settings.
Technical feasibility todayclaude-sonnet-53/5Digital logbook and CMMS/SCADA products with automated data capture exist and are deployed in some industrial facilities, but many boiler operations still rely on manual paper or basic digital logs without full automation, and malfunction narrative descriptions require human judgment.

Observe and interpret readings on gauges, meters, and charts registering various aspects of boiler operation to ensure that boilers are operating properly.

38

CI 34–43 · exposure 42 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Most boiler operations remain in traditional industrial settings (manufacturing, utilities, facilities management) with slow digital adoption; while modern plants use SCADA, replacement of human interpretation is limited and operator presence is often legally mandated.
Sector adoption velocityclaude-sonnet-52/5Industrial/manufacturing and facilities sectors are slower adopters of AI compared to information/professional services, though predictive maintenance and IoT sensor adoption is growing steadily in larger industrial operations.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-powered monitoring dashboards, predictive alerts, and anomaly detection substantially enhance operator productivity by highlighting abnormalities and trends across multiple gauges simultaneously, allowing faster diagnosis and intervention while the operator maintains oversight and control.
Augmentation potentialclaude-sonnet-54/5AI-enabled monitoring dashboards, anomaly detection, and predictive alerts significantly help operators track boiler conditions more efficiently and catch issues earlier while the licensed operator remains responsible for oversight and response.
Task automatabilityclaude-haiku-4-5-202510012/5Modern SCADA and monitoring systems can automatically log and alert on gauge readings, but interpretation of anomalies and real-time decision-making about boiler state typically requires human judgment and domain knowledge that current AI cannot reliably replicate without expert supervision.
Task automatabilityclaude-sonnet-53/5Sensor readings can be digitized and monitored via SCADA/IoT systems with automated alerts, but full end-to-end automation requires integration with legacy analog equipment and physical presence for verification and response, limiting full replacement today.
Adoption barriersclaude-haiku-4-5-202510014/5Boiler operation in many jurisdictions requires a licensed, certified operator legally responsible for safe operation; liability and safety-critical nature of the task create strong regulatory and organizational friction against full automation without human sign-off.
Adoption barriersclaude-sonnet-54/5Many jurisdictions require licensed stationary engineers/boiler operators to be present and legally responsible for boiler safety, creating strong regulatory barriers to full automation of this monitoring task.
Cost vs. human wageclaude-haiku-4-5-202510013/5Industrial monitoring automation has moderate upfront cost but low per-reading inference cost; however, integration, calibration, and human oversight requirements keep total cost-of-ownership comparable to experienced operator wages in many facilities.
Cost vs. human wageclaude-sonnet-53/5Sensor and monitoring system installation plus software costs can be significant upfront, but ongoing monitoring is cheap; overall cost is roughly comparable to human labor once amortized, especially for smaller facilities with older equipment.
Technical feasibility todayclaude-haiku-4-5-202510013/5Automated monitoring systems exist in industrial settings and can detect out-of-range values, but fully autonomous interpretation of complex boiler operation status—including trending, cross-gauge diagnosis, and failure prediction—remains partially manual in most production environments.
Technical feasibility todayclaude-sonnet-53/5Industrial monitoring products (SCADA, DCS, predictive maintenance platforms) reliably track boiler parameters in many facilities, but many boiler operations still rely on operators doing physical rounds and reading analog gauges not yet digitized.

Monitor boiler water, chemical, and fuel levels, and make adjustments to maintain required levels.

28

CI 25–30 · exposure 34 · augmentation 75 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Boiler operation is concentrated in older industrial, utility, and building-management sectors with slower digital transformation. While large facilities may adopt monitoring systems, displacement of operators is minimal due to regulatory requirements and the distributed, fragmented nature of small-to-medium boiler installations.
Sector adoption velocityclaude-sonnet-52/5Industrial and facilities/utilities sectors adopt automation slowly due to capital intensity, safety-critical regulation, and legacy equipment, with full autonomous replacement rare.
Augmentation potentialclaude-haiku-4-5-202510014/5Automated monitoring dashboards, predictive alerts for chemical imbalances, and real-time level visualization significantly assist operators in making faster, more informed adjustments. AI-driven analytics can flag anomalies before they require corrective action, raising operator productivity while the engineer remains in full control.
Augmentation potentialclaude-sonnet-54/5Automated sensors, alarms, and control system dashboards significantly assist operators in tracking water/chemical/fuel levels and flagging anomalies, improving efficiency and safety while humans remain responsible for oversight.
Task automatabilityclaude-haiku-4-5-202510012/5Modern boiler systems have automated sensors and control systems that can monitor levels and make some adjustments, but the task requires judgment about chemical balance, fuel efficiency, and system safety that demands human oversight. Full end-to-end automation with 50% time savings at equal quality is not achievable today without substantial human validation.
Task automatabilityclaude-sonnet-52/5Sensor-based monitoring can be automated via SCADA/control systems, but the full task including judgment-based adjustments and physical valve/dosing actions still requires human presence and intervention in most facilities.
Adoption barriersclaude-haiku-4-5-202510015/5Federal and state regulations (ASME Boiler and Pressure Vessel Code, EPA) typically require a licensed stationary engineer or certified operator to be responsible for boiler operation and safety certification. Liability and safety-critical nature of the task create hard legal barriers to full automation without a licensed human sign-off.
Adoption barriersclaude-sonnet-54/5Boiler operation is often subject to licensing requirements and safety regulations (e.g., ASME codes, state boiler licensing laws) requiring certified operators to be present or responsible for oversight.
Cost vs. human wageclaude-haiku-4-5-202510012/5Initial sensor and monitoring system installation is capital-intensive, and integration with existing boiler infrastructure adds significant cost. Ongoing maintenance, calibration, and the continued need for human oversight mean total cost approaches or exceeds the loaded wage of a stationary engineer.
Cost vs. human wageclaude-sonnet-52/5Automated control systems have high upfront capital and integration costs relative to a single operator's wage, especially for smaller facilities, though large plants may already have amortized such systems.
Technical feasibility todayclaude-haiku-4-5-202510013/5SCADA systems and IoT sensors can perform real-time monitoring and log data reliably, and some systems include automated adjustment capabilities for routine level maintenance. However, deployed products typically require human operators to interpret alerts, authorize major adjustments, and troubleshoot anomalies, so reliability is partial rather than complete.
Technical feasibility todayclaude-sonnet-53/5Automated control systems (DCS/SCADA with automatic feedwater and chemical dosing) are mature and widely deployed in industrial boiler plants, though many smaller or older facilities still rely on manual monitoring and adjustment.

Monitor and inspect equipment, computer terminals, switches, valves, gauges, alarms, safety devices, and meters to detect leaks or malfunctions and to ensure that equipment is operating efficiently and safely.

28

CI 25–30 · exposure 30 · augmentation 63 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Stationary engineering occurs in traditional industrial, manufacturing, and facilities sectors that digitize slowly. While large facilities adopt SCADA systems, meaningful AI-driven replacement of inspection tasks remains rare in production, with most adoption limited to pilots or supplementary monitoring.
Sector adoption velocityclaude-sonnet-52/5Industrial/utility sectors adopt monitoring technology slowly due to legacy infrastructure, safety regulation, and capital cycles, with pilots more common than widespread agentic automation.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-assisted dashboards, anomaly alerts, and predictive maintenance tools can help operators prioritize equipment checks and interpret complex sensor patterns, raising efficiency on data-heavy aspects. However, the physical inspection and contextual judgment components limit the scope of human productivity transformation.
Augmentation potentialclaude-sonnet-54/5AI-enabled sensors, dashboards, and predictive alerts significantly help operators detect anomalies faster and prioritize inspections, meaningfully boosting productivity while humans remain responsible for final checks.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can monitor some sensor data and detect threshold violations programmatically, the task requires physical inspection of equipment, identification of subtle leaks or wear patterns, and contextual judgment about operational safety—capabilities that current AI systems lack end-to-end. Visual inspection of physical systems remains largely manual.
Task automatabilityclaude-sonnet-52/5Sensor monitoring and anomaly detection can be partially automated with IoT/SCADA systems, but physical inspection, leak detection by smell/sound, and safety judgment calls still require human presence and cannot fully meet the 50% time-saving bar end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Safety-critical infrastructure is heavily regulated (ASME, OSHA, EPA rules); operators must be licensed and certified to perform and sign off on inspections. Liability for equipment failure or safety incidents creates strong legal and insurance barriers to full automation without human oversight.
Adoption barriersclaude-sonnet-54/5Boiler operation is often subject to licensing and regulatory requirements (e.g., certified boiler operators), and safety-critical inspections typically require a legally responsible human on-site.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-based monitoring systems (sensors, inference, integration) represent significant capital investment and ongoing costs, while human operators perform continuous visual and auditory inspection at relatively modest wages. The all-in cost per shift of monitoring typically favors human presence today.
Cost vs. human wageclaude-sonnet-52/5Sensor networks and monitoring software have upfront and maintenance costs comparable to or exceeding a portion of operator wages, and physical inspection still requires paid staff, limiting cost savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5Deployed IoT and SCADA systems can monitor digital gauges and alarms, but reliable autonomous detection of leaks, valve malfunctions, and equipment degradation across diverse physical equipment remains in early stages. Products exist for narrowband monitoring but not for the full spectrum of physical inspection and anomaly detection.
Technical feasibility todayclaude-sonnet-53/5Building automation systems, SCADA, and predictive maintenance sensors are deployed in many industrial plants today, but they augment rather than replace the physical walk-through inspections operators perform.

Activate valves to maintain required amounts of water in boilers, to adjust supplies of combustion air, and to control the flow of fuel into burners.

26

CI 18–34 · exposure 30 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Boiler operation remains concentrated in manufacturing, utilities, and institutional facilities with low digital adoption velocity and strong regulatory conservatism. Most plants continue to employ human operators due to regulatory requirements and the high cost of retrofitting legacy systems with new automation.
Sector adoption velocityclaude-sonnet-52/5Industrial/facilities sectors employing boiler operators adopt automation slowly due to legacy equipment, safety regulation, and capital cycles; full autonomous control adoption is limited compared to fast-moving digital sectors.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by providing real-time alerts, predictive maintenance recommendations, and automated logging of valve positions and sensor readings, improving operator awareness and reducing manual monitoring burden, though the operator remains responsible for all control decisions.
Augmentation potentialclaude-sonnet-53/5Modern control systems and sensors assist operators by providing real-time data and automated adjustments, improving efficiency and safety, though the operator still monitors and intervenes.
Task automatabilityclaude-haiku-4-5-202510012/5While valve activation could theoretically be automated via industrial controls, the task requires real-time monitoring of multiple interdependent systems and judgment calls based on system state, pressure readings, and safety conditions that current AI agents handle poorly in high-stakes environments. Only discrete, pre-programmed valve sequences can be reliably automated today; complex, adaptive control still requires human oversight.
Task automatabilityclaude-sonnet-52/5Physical valve manipulation and real-time control of boiler water/air/fuel is possible via existing SCADA/DCS automation, but this task as stated describes a hands-on operator action requiring physical presence and judgment, not a pure software/cognitive task an LLM-based AI could perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510014/5Boiler operations are subject to ASME codes, state and local steam equipment regulations, and insurance requirements that often mandate a licensed stationary engineer or operator physically present and responsible for safe operation. Liability and safety certification create hard legal barriers to full automation.
Adoption barriersclaude-sonnet-55/5Boiler operation is subject to strict licensing requirements (stationary engineer licenses) and safety regulations mandating qualified personnel to operate or supervise combustion equipment due to explosion/safety risk.
Cost vs. human wageclaude-haiku-4-5-202510012/5Retrofitting a boiler system with advanced AI-driven control infrastructure is capital-intensive and requires ongoing integration and safety validation costs; a stationary engineer's wage is modest, making replacement economically unfavorable compared to incremental automation of individual steps.
Cost vs. human wageclaude-sonnet-53/5Control automation hardware/software has high upfront integration and maintenance costs comparable to a technician's wage over time, though at scale automated systems can be cheaper per operating hour than continuous human monitoring.
Technical feasibility todayclaude-haiku-4-5-202510012/5SCADA and PLC systems exist for boiler automation, but these are specialized industrial control systems, not AI products. AI agents lack demonstrated capability to safely manage the full scope of water level, air supply, and fuel flow control with the reliability required in production boiler operations where failures have safety consequences.
Technical feasibility todayclaude-sonnet-53/5Automated boiler control systems (PLCs, DCS) reliably regulate water level, combustion air, and fuel flow in many industrial plants today, but full replacement of the human operator's oversight and intervention role is not standard in most facilities requiring licensed operators.

Check the air quality of ventilation systems and make adjustments to ensure compliance with mandated safety codes.

25

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI monitoring tools is still emerging in industrial facilities. While some large organizations pilot sensor networks and analytics, meaningful displacement of the inspection and adjustment functions remains rare; most facilities rely on scheduled manual checks by trained operators.
Sector adoption velocityclaude-sonnet-52/5Facilities management and industrial plant operations are traditionally slow adopters of AI compared to information-sector jobs, with physical infrastructure and legacy systems slowing digitization.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can augment this task through real-time air quality dashboards, automated alerts when thresholds are approached, and recommended adjustments, helping operators respond faster and more consistently. However, the augmentation is limited to enhancing monitoring and decision-support rather than transforming overall productivity.
Augmentation potentialclaude-sonnet-54/5AI-enabled sensors and predictive analytics can meaningfully help operators monitor air quality trends and flag deviations, improving efficiency while the licensed human still performs adjustments and compliance sign-off.
Task automatabilityclaude-haiku-4-5-202510012/5While AI systems can analyze air quality sensor data and flag deviations from standards, this task requires physical adjustments to ventilation systems in response to real-time conditions, which remains beyond current automation. Meaningful parts (data monitoring, threshold detection) could be automated, but the full end-to-end task including making physical adjustments and ensuring compliance verification falls short of the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5Sensor-based monitoring can be automated and AI can flag anomalies, but physical inspection, adjustment of equipment, and final compliance judgment still require a human on-site.'
Adoption barriersclaude-haiku-4-5-202510014/5Significant regulatory and liability barriers exist: safety codes typically require a licensed stationary engineer to verify and sign off on compliance; OSHA and local codes often mandate human inspection and accountability for air quality standards. Legal liability for unsafe conditions creates a hard requirement for human authority.
Adoption barriersclaude-sonnet-54/5Boiler and ventilation systems are subject to safety codes often requiring licensed stationary engineers or certified operators to inspect and sign off on compliance, creating strong regulatory and liability barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of implementing continuous AI monitoring, sensor infrastructure, integration with building systems, and ongoing human oversight for compliance sign-off approaches or exceeds the loaded wage of a stationary engineer. Savings are marginal without full automation.
Cost vs. human wageclaude-sonnet-52/5Sensor networks and monitoring software have upfront and maintenance costs comparable to or sometimes exceeding the marginal cost of a technician periodically checking systems, especially amortized in smaller facilities.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI-based monitoring systems exist and can process air quality data in real time, but no deployed product reliably performs the full task of checking air quality *and* making adjustments autonomously. Human operators remain necessary for physical interventions and judgment calls on complex compliance scenarios; current tools are assistive rather than autonomous.
Technical feasibility todayclaude-sonnet-52/5Building management systems with AI-driven air quality sensors exist, but they typically alert rather than autonomously adjust equipment to meet code, and human verification remains standard practice.

Develop operation, safety, and maintenance procedures or assist in their development.

24

CI 23–25 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Industrial boiler operations remain heavily regulated and physically grounded; adoption of AI for procedure development is minimal outside large facilities with dedicated safety teams, and regulatory conservatism in the energy/utilities sector slows experimental adoption of AI-generated compliance documentation.
Sector adoption velocityclaude-sonnet-52/5Industrial/facilities engineering sectors are slow adopters of AI for safety-critical documentation, with adoption concentrated in pilot programs rather than widespread production use.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by generating initial drafts, surfacing relevant regulatory requirements, and identifying procedure gaps through template matching, meaningfully reducing the human engineer's time on rote sections; however, the core work of validating safety, site-specificity, and legal compliance remains human-driven.
Augmentation potentialclaude-sonnet-54/5AI tools can meaningfully speed up drafting, formatting, and referencing standards for procedure documents, letting engineers focus on technical validation and customization.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can generate draft procedures and checklists based on regulatory standards and equipment specs, the task requires deep domain expertise, equipment-specific knowledge, and understanding of site-specific constraints that currently available AI systems cannot reliably synthesize at production quality. Human engineers must validate and refine outputs for liability and safety compliance, limiting time savings below the 50% threshold.
Task automatabilityclaude-sonnet-52/5Drafting procedures requires deep plant-specific technical knowledge, regulatory context, and equipment familiarity that AI cannot fully substitute; AI can assist drafting but not autonomously produce validated procedures end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5ASME, insurance, and regulatory standards typically require that operational procedures for boiler systems be developed and signed by or under the direct supervision of licensed Professional Engineers or certified operators, creating a hard legal barrier to full automation or unsupervised AI generation.
Adoption barriersclaude-sonnet-54/5Safety and maintenance procedures for boilers are subject to regulatory codes (e.g., ASME, OSHA) and typically require sign-off by licensed engineers or qualified personnel, creating strong liability and compliance barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI can reduce drafting time for routine sections, but safety procedure development requires licensed engineers for sign-off and validation. The integrated cost of AI generation plus mandatory human expert review and liability handling is comparable to or exceeds direct human authorship for complex systems.
Cost vs. human wageclaude-sonnet-52/5AI drafting assistance is cheap per document, but the human engineering review, validation against codes, and liability sign-off dominate cost, keeping overall savings modest.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI can draft procedure templates and provide regulatory guidance via document analysis, but no deployed products reliably generate complete, validated operational procedures for industrial boiler systems without substantial human expert review. Current capabilities are narrow and error rates on safety-critical procedure generation are too high for production use without extensive oversight.
Technical feasibility todayclaude-sonnet-52/5LLMs can generate draft SOPs from templates or manuals, but no deployed product reliably produces site-specific, code-compliant safety/maintenance procedures for boiler operations without heavy human review.

Operate or tend stationary engines, boilers, and auxiliary equipment, such as pumps, compressors, or air-conditioning equipment, to supply and maintain steam or heat for buildings, marine vessels, or pneumatic tools.

23

CI 20–25 · exposure 25 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption is slow in this sector. While some large facilities pilot predictive maintenance, most small-to-medium boiler operations remain manually staffed and resistant to change given safety criticality, regulatory requirements, and the aging workforce's institutional knowledge.
Sector adoption velocityclaude-sonnet-52/5Facilities and industrial/marine sectors are slow adopters of full automation for physical plant operations, though building management systems with sensor analytics are gradually increasing in use.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist operators through predictive maintenance alerts, real-time performance dashboards, and anomaly detection that reduces cognitive load and reaction time. However, the augmentation is partial—human operators remain essential for judgment, manual adjustments, and emergency response.
Augmentation potentialclaude-sonnet-53/5AI-enabled monitoring, predictive maintenance alerts, and anomaly detection can meaningfully assist operators in tracking system performance and anticipating failures, improving efficiency while humans remain responsible for control actions.
Task automatabilityclaude-haiku-4-5-202510012/5While AI monitoring systems can detect anomalies and log data, this task requires real-time physical intervention (valve adjustments, emergency shutdowns, manual repairs) that AI cannot perform. Current systems can support decision-making but cannot autonomously operate equipment or provide the 50% time savings threshold needed for full automation.
Task automatabilityclaude-sonnet-52/5This is a hands-on physical monitoring and control task requiring presence at equipment, manual adjustments, and physical intervention that current AI cannot perform end-to-end; sensor-based monitoring can be partially automated but the full operating role remains human.the task.
Adoption barriersclaude-haiku-4-5-202510014/5Strong regulatory and safety barriers exist: boiler operations are governed by state and federal codes requiring licensed operators; insurance and liability frameworks mandate human sign-off for safety-critical decisions; and ASME standards typically require a credentialed operator on-site for active systems.
Adoption barriersclaude-sonnet-55/5Boiler operation is heavily regulated, often requiring licensed stationary engineers by law, with significant safety and liability risk (explosion, injury) mandating human accountability and legal sign-off.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI monitoring and predictive maintenance systems require significant upfront infrastructure investment and continuous oversight labor. The cost per task-equivalent remains comparable to or exceeds the loaded wage of a skilled stationary engineer given integration, maintenance, and liability overhead.
Cost vs. human wageclaude-sonnet-52/5Automation/monitoring software is cheap to run, but full task substitution still requires human presence and physical intervention plus certified oversight, keeping all-in costs comparable to or only modestly below human labor.
Technical feasibility todayclaude-haiku-4-5-202510012/5Monitoring software exists and some facilities use automated alarm systems, but no deployed AI products reliably operate stationary engines and boilers end-to-end. Existing systems are narrow (alerting only) and require constant human oversight for safety-critical equipment.
Technical feasibility todayclaude-sonnet-52/5Building automation systems and SCADA/IoT monitoring products exist and are deployed for boiler telemetry and alerts, but no product autonomously operates and tends boilers/auxiliary equipment without a licensed human present.

Adjust controls and/or valves on equipment to provide power, and to regulate and set operations of system or industrial processes.

23

CI 20–25 · exposure 25 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Industrial sectors operating boilers are traditionally slow to adopt cutting-edge automation due to safety criticality, capital intensity, regulatory conservatism, and the long operational life of legacy equipment; adoption remains largely incremental and localized to advanced facilities.
Sector adoption velocityclaude-sonnet-52/5Industrial/manufacturing sectors adopt automation for control processes but do so slowly relative to information sectors, with significant capital and safety certification requirements slowing AI-specific adoption for this task.
Augmentation potentialclaude-haiku-4-5-202510013/5Modern SCADA dashboards and sensor networks assist operators by providing real-time data visualization and alerting, improving situational awareness; however, the core task of manual valve adjustment still rests with the human, limiting the transformative upside of augmentation.
Augmentation potentialclaude-sonnet-53/5AI-based monitoring and predictive analytics can assist operators by flagging anomalies and suggesting adjustments, improving efficiency, though the operator still manually executes and verifies actions.
Task automatabilityclaude-haiku-4-5-202510012/5While routine monitoring and simple setpoint adjustments might be partially automated, the task requires real-time physical adjustment of valves and controls in response to complex, variable industrial processes—a domain where current AI lacks robust end-to-end autonomy, safe hardware integration, and the contextual judgment needed for safe operation.
Task automatabilityclaude-sonnet-52/5Some routine control adjustments can be handled by existing SCADA/automation systems, but the task as described requires real-time physical presence, sensory judgment, and response to equipment anomalies that current AI cannot fully replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Stationary engineers and boiler operators are licensed professionals in most jurisdictions; federal and state regulations (ASME Boiler and Pressure Vessel Code, state licensure) typically mandate that a licensed human must operate and sign off on boiler system control, creating hard legal barriers to full automation.
Adoption barriersclaude-sonnet-54/5Stationary engineers and boiler operators often require licensure, and safety regulations mandate human oversight of high-pressure systems, creating strong legal and liability barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current automation for these tasks (industrial control systems, sensors, integration) is expensive to install and maintain relative to the wages of a single operator, especially when accounting for safety certification, redundancy, and fail-safe requirements in safety-critical boiler operations.
Cost vs. human wageclaude-sonnet-52/5While automated control systems are cost-effective for routine adjustments, the sensors, safety redundancy, and integration needed to replace a human operator's judgment across all scenarios make all-in AI costs comparable to or higher than a single operator's wage in many facilities.
Technical feasibility todayclaude-haiku-4-5-202510012/5Although some SCADA and PLC systems incorporate basic automation for setpoint management, they are narrow supervisory tools requiring human oversight; no deployed product reliably performs independent physical valve adjustment and control sequencing across diverse boiler systems without human validation and intervention.
Technical feasibility todayclaude-sonnet-52/5Industrial control systems and PLCs already automate many setpoint adjustments, but fully autonomous AI-driven adjustment without human oversight in boiler/plant operations is not standard deployed practice due to safety-critical constraints.

Analyze problems and take appropriate action to ensure continuous and reliable operation of equipment and systems.

21

CI 16–25 · exposure 17 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Industrial facilities have adopted monitoring dashboards and predictive maintenance tools, but remain cautious about autonomous decision-making in safety-critical operations. Digitization exists but adoption of AI-driven corrective action remains slow and pilot-focused, particularly in smaller and older plants.
Sector adoption velocityclaude-sonnet-52/5Industrial/facilities sectors adopt monitoring tech gradually; this is a physically-embedded, safety-critical role with slower AI integration than office-based professional work.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by ingesting sensor data, identifying patterns, and recommending diagnostics, raising an operator's ability to respond quickly. However, the human must still interpret context, authorize actions, and take responsibility—a classic augmentation scenario where AI enhances rather than replaces judgment.
Augmentation potentialclaude-sonnet-54/5AI-driven sensors, predictive analytics, and diagnostic tools significantly help operators detect problems earlier and prioritize responses, meaningfully boosting productivity while the human remains in control.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in monitoring equipment data and flagging anomalies, the task requires judgment calls about appropriate corrective actions in complex, varied operational contexts that demand human expertise and responsibility. End-to-end automation with 50% time savings would require AI to reliably diagnose and execute repairs across diverse equipment—a capability not yet demonstrated in production stationary engineering environments.
Task automatabilityclaude-sonnet-51/5This requires physical inspection, hands-on diagnosis, and manual corrective actions on boilers and mechanical systems that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Stationary engineers must be licensed in most jurisdictions and legally responsible for equipment safety and code compliance. Liability for equipment failure, worker safety, and regulatory violations creates strong disincentives to full automation without certified human sign-off, and the consequence of errors (explosions, environmental damage) is severe.
Adoption barriersclaude-sonnet-54/5Boiler operation often requires licensed operators by law/safety regulation, and equipment failures carry high safety and liability risk, creating strong barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI monitoring and diagnostic tools require significant integration with existing SCADA/control systems, skilled human oversight, and liability coverage. The all-in cost remains comparable to or higher than paying an experienced boiler operator, especially when accounting for the safety-critical nature of errors.
Cost vs. human wageclaude-sonnet-52/5Sensor/monitoring systems add cost on top of still-required human labor for diagnosis and physical intervention, so total cost is not clearly cheaper than a human operator alone.
Technical feasibility todayclaude-haiku-4-5-202510012/5Monitoring systems with AI alerts exist in industrial settings, but no deployed product reliably performs autonomous problem diagnosis and corrective action selection for stationary equipment at scale. Most production systems flag issues for human operators rather than independently determining and executing remediation.
Technical feasibility todayclaude-sonnet-52/5Predictive maintenance and monitoring software exist and are deployed for anomaly detection, but the actual troubleshooting and corrective action still requires a human operator on-site.

Contact equipment manufacturers or appropriate specialists when necessary to resolve equipment problems.

21

CI 14–28 · exposure 17 · augmentation 38 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This sector comprises small-to-medium industrial sites with legacy equipment and slower digital transformation; adoption of AI for critical equipment-management decisions remains minimal.
Sector adoption velocityclaude-sonnet-51/5Stationary engineering and boiler operation is a low-digitization, physical-plant occupation with minimal AI agent adoption in production environments.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by summarizing troubleshooting logs or suggesting potential contacts, but the core judgment and communication require the operator's expertise and responsibility, limiting augmentation impact.
Augmentation potentialclaude-sonnet-53/5AI can help draft problem descriptions, search manuals, or suggest likely specialists, offering moderate assistance while the human still makes contact and judgment calls.
Task automatabilityclaude-haiku-4-5-202510011/5This task inherently requires human judgment to diagnose complex equipment issues and establish direct communication with manufacturers or specialists—capabilities that demand contextual reasoning and interpersonal negotiation beyond current AI automation.
Task automatabilityclaude-sonnet-52/5Involves judgment about which specialist to contact, describing complex equipment problems, and coordinating human relationships; AI could draft communications but the diagnostic decision-making and outreach remain human-driven.9
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: stationary engineers and boiler operators are licensed professionals with regulatory accountability for equipment safety, and manufacturers often require signed communication from qualified operators; liability for incorrect contact decisions is asymmetric.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement for making a phone call, but liability and safety concerns around boiler operation create organizational reluctance to delegate this judgment call to AI.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI could assist in identifying contacts and drafting initial communications, but the core task—deciding when and whom to contact—still requires human expertise; savings are modest relative to operator wages.
Cost vs. human wageclaude-sonnet-52/5Human engineers already handle this quickly as part of routine duties; AI would add integration and oversight costs without clear savings for an infrequent, judgment-heavy task.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can draft communication templates or retrieve manufacturer contact information, no deployed system reliably diagnoses which specialist to contact or negotiates complex technical issues end-to-end without human oversight.
Technical feasibility todayclaude-sonnet-52/5AI chat/email drafting tools exist, but no deployed product autonomously identifies equipment problems and initiates specialist contact in industrial boiler settings today.

Investigate and report on accidents.

17

CI 14–20 · exposure 16 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI in accident investigation is slow because investigations are infrequent, high-stakes events with regulatory scrutiny. Even digitization-forward sectors move cautiously; most organizations still rely on human expertise and documentation rather than AI-driven investigation.
Sector adoption velocityclaude-sonnet-51/5Stationary engineering and boiler operation is a low-digitization, physical-plant sector with minimal AI agent adoption for safety investigations.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by organizing accident data, flagging similar past incidents, drafting preliminary reports, and highlighting anomalies in sensor readings, thereby helping human investigators work more efficiently. However, the assistant role is secondary to human judgment on root cause.
Augmentation potentialclaude-sonnet-53/5AI tools can help organize findings, draft reports, and cross-reference safety codes, meaningfully aiding the reporting portion of the task.
Task automatabilityclaude-haiku-4-5-202510012/5Accident investigation requires on-site inspection, physical evidence collection, and contextual judgment that current AI cannot perform autonomously. While AI can assist with report drafting and data analysis, the investigative core—observing damage patterns, interviewing witnesses, determining causation in complex mechanical failures—remains fundamentally human work.
Task automatabilityclaude-sonnet-52/5Investigating a physical accident requires on-site inspection, evidence gathering, and interviewing witnesses, which AI cannot perform; AI can assist with drafting the report portion but not the core investigative work.4}
Adoption barriersclaude-haiku-4-5-202510014/5Accident investigation is often subject to regulatory oversight (OSHA, insurance requirements, legal liability). Many jurisdictions require a qualified engineer or inspector to formally investigate workplace incidents, and liability for incorrect findings creates strong organizational and legal barriers to full automation.
Adoption barriersclaude-sonnet-54/5Accident investigations often trigger regulatory (OSHA) and liability requirements demanding a qualified human's sign-off and judgment, limiting substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5Current AI cannot replace the investigator; it can only draft reports from human-gathered evidence. The cost of AI oversight, fact-checking, and investigation leadership would exceed the savings from report automation, leaving human labor as the dominant cost.
Cost vs. human wageclaude-sonnet-52/5Human investigators are still required to physically assess sites and equipment, so AI only reduces cost for the report-writing portion, not the full task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed products perform end-to-end accident investigation reliably. AI tools exist for document processing and incident logging, but the investigative judgment and synthesis needed to determine root causes in boiler/machinery accidents are not reliably automated in production systems.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously conducts physical accident investigations for equipment failures; this remains a research-stage capability at best.

Test electrical systems to determine voltages, using voltage meters.

16

CI 14–19 · exposure 20 · augmentation 38 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Stationary engineers work in industrial/utility sectors with slower digital transformation; physical presence on-site is mandatory, and organizational inertia around safety-critical testing tasks is strong. Adoption of automation in this role remains minimal.
Sector adoption velocityclaude-sonnet-51/5Stationary engineering and boiler operation is a physical, hands-on trade with low digitization and minimal AI/agent adoption in production settings.
Augmentation potentialclaude-haiku-4-5-202510013/5Digital multimeters and data logging software already assist technicians in recording and trending voltage measurements, and AI could enhance this by flagging anomalies or predicting maintenance needs. However, the core testing act remains human-dependent.
Augmentation potentialclaude-sonnet-52/5AI could assist with logging readings, trend analysis, or flagging anomalies from recorded voltage data, but it does not meaningfully assist the core physical act of testing.
Task automatabilityclaude-haiku-4-5-202510012/5While AI-driven vision systems could theoretically identify meter readings from images, the task requires hands-on placement of probes at precise circuit points and interpretation of safety-critical measurements in real-world industrial contexts. Current systems cannot reliably perform the full end-to-end task of testing multiple electrical points with proper safety protocols and achieving 50% time savings.
Task automatabilityclaude-sonnet-52/5This requires physical presence to attach meters to equipment and interpret readings in context, which current AI systems cannot perform end-to-end without robotic hardware.forces they lack today.dad occupational deployment.rically.hearing.mediately.appropriately.diliges.arily.ously.uate.tually.esely.ributely.rmally.antly.rectly.mently.pletely.iciently.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: electrical work on live systems requires licensed operators in most jurisdictions, liability for incorrect voltage readings is high, and regulatory codes mandate human responsibility for testing decisions. Safety-critical nature creates legal barriers to full automation.
Adoption barriersclaude-sonnet-54/5Electrical safety testing on industrial equipment typically requires certified/licensed personnel due to safety and liability concerns, creating a hard barrier to non-human performance.
Cost vs. human wageclaude-haiku-4-5-202510011/5The hardware (robotic arms, vision systems, safety-certified integration) required for autonomous electrical testing far exceeds the cost of a skilled technician performing the task. Oversight and recalibration would add significant ongoing expense.
Cost vs. human wageclaude-sonnet-51/5Without robotic manipulation capability, AI cannot substitute for the human physically performing this test, so there is no viable AI cost comparison; a human operator remains necessary.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed products reliably automate this task end-to-end. Computer vision can read analog/digital meters from images with moderate accuracy, but integrating probe placement, circuit access, and safety verification into a functioning industrial automation system is not mature or available at scale.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product autonomously performs physical voltage testing with handheld meters on industrial boiler/electrical systems; this remains a manual, hands-on inspection task.

Fire coal furnaces by hand or with stokers and gas- or oil-fed boilers, using automatic gas feeds or oil pumps.

15

CI 5–25 · exposure 13 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Industrial boiler operations remain in legacy, regulated sectors with strong union presence and slow digitization. While some larger facilities have adopted automated controls, most plants still rely on human operators; adoption of AI-driven agents is minimal and confined to monitoring, not autonomous operation.
Sector adoption velocityclaude-sonnet-51/5Industrial/utility/facilities sectors operating boilers are slow to adopt AI-driven physical automation; existing control systems are traditional SCADA/PLC-based, not AI agents.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-powered monitoring dashboards and predictive maintenance alerts can assist operators by flagging anomalies and optimizing fuel flow, raising situational awareness and efficiency. However, the task remains fundamentally manual and safety-critical, limiting the transformative potential of augmentation.
Augmentation potentialclaude-sonnet-52/5AI-based predictive maintenance and combustion optimization software can assist engineers in monitoring efficiency and detecting anomalies, but it doesn't materially change the hands-on firing/fueling task itself.
Task automatabilityclaude-haiku-4-5-202510012/5While modern boiler systems use automatic gas feeds and oil pumps, this task requires real-time monitoring, pressure regulation, and fault diagnosis in a physical environment with safety-critical constraints. Current AI cannot reliably operate physical stokers or respond to unexpected furnace conditions without human oversight, preventing the ≥50% time-saving threshold.
Task automatabilityclaude-sonnet-51/5This is a hands-on physical task requiring in-person operation and monitoring of combustion equipment; no AI system can physically fire furnaces or manage fuel feed mechanisms end-to-end today.
Adoption barriersclaude-haiku-4-5-202510014/5Boiler operation is heavily regulated under ASME codes and state licensing requirements; stationary engineers must be licensed, and an authorized human operator must be on-site for safety certification and liability. Legal and regulatory mandates create hard barriers to full automation.
Adoption barriersclaude-sonnet-54/5Boiler operation is often subject to licensing requirements and safety regulations (pressure vessel codes, OSHA), and errors can cause explosions or injury, creating strong liability and regulatory barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Retrofit automation systems and AI-driven controls are capital-intensive and require ongoing integration and tuning. The cost per task equivalent remains comparable to or exceeds the loaded wage of a stationary engineer, especially when accounting for safety compliance and liability.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical task, so cost comparison is moot; any solution would require expensive robotics/control retrofits exceeding human wage costs by far.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some boiler control systems have automated ignition and fuel feed regulation, but deployed products handle only routine conditions; complex scenarios (fouling, pressure surges, safety shutdowns) still require human operators. No production system fully autonomously fires and maintains furnaces end-to-end.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs this physical firing/fueling task; existing automation in this space is decades-old industrial control systems, not AI-driven autonomous operation.

Test boiler water quality or arrange for testing and take necessary corrective action, such as adding chemicals to prevent corrosion and harmful deposits.

15

CI 5–25 · exposure 17 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Boiler rooms are legacy, capital-intensive, and risk-averse sectors with long equipment lifecycles and strong regulatory oversight. Adoption of AI agents in this domain remains minimal; organizations rely on certified human operators and periodic third-party lab testing rather than autonomous systems.
Sector adoption velocityclaude-sonnet-52/5Industrial/facilities maintenance sectors adopt automation more slowly than knowledge work; while SCADA/IoT monitoring is spreading, full displacement of manual testing and corrective action remains limited.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by analyzing historical water-quality trends, flagging anomalies, and recommending corrective actions (e.g., dosing amounts) for the operator's review. However, the operator remains essential for field sampling, physical intervention, and regulatory sign-off, limiting transformative augmentation.
Augmentation potentialclaude-sonnet-54/5Automated sensors and dosing control systems significantly assist operators by providing continuous water quality data and dosing recommendations, improving efficiency while the operator remains responsible for verification and corrective decisions.
Task automatabilityclaude-haiku-4-5-202510011/5Testing water quality and chemical adjustment require real-time sampling, interpretation of multi-variable chemical results, and physical manipulation of treatment systems in response to conditions. Current AI cannot autonomously perform the hands-on sampling, chemical analysis, or precise real-world intervention needed.
Task automatabilityclaude-sonnet-52/5Physical sampling, chemical dosing, and equipment adjustment require hands-on presence; while sensors can automate continuous monitoring, the full task including physical testing and corrective chemical addition cannot be fully automated end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Boiler operation is heavily regulated by ASME and state/local codes; stationary engineers must be licensed, and water treatment is legally mandated. Liability for boiler failure and hazards (pressure, corrosion, explosions) places automation under strict regulatory scrutiny; a human operator must legally verify and sign off on water quality and corrective actions.
Adoption barriersclaude-sonnet-54/5Boiler operation is often subject to safety regulations and licensing requirements, and improper water treatment can cause costly equipment damage or safety incidents, creating strong liability and compliance barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The task requires specialized chemical testing equipment and real-time physical intervention; autonomous systems would need redundant sensing, calibration, and safety certification. The capital and oversight costs far exceed the hourly wage of a trained boiler operator who performs this as routine work.
Cost vs. human wageclaude-sonnet-52/5Automated water treatment/monitoring systems have meaningful upfront capital and maintenance costs comparable to or exceeding the marginal cost of a technician performing periodic tests and adjustments, especially at smaller facilities.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist in interpreting water-quality data (e.g., recommending corrective actions based on test results), no deployed product autonomously tests boiler water or adjusts chemical treatment in situ. Deployed systems are narrow laboratory analytics, not field-deployed autonomous boiler maintenance.
Technical feasibility todayclaude-sonnet-52/5Automated water chemistry monitoring systems exist in industrial boiler plants, but they typically alert or recommend rather than autonomously executing corrective chemical treatment reliably without human oversight.

Operate mechanical hoppers and provide assistance in their adjustment and repair.

9

CI 5–14 · exposure 8 · augmentation 25 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of autonomous hopper operation is nearly non-existent in production settings. The industrial/manufacturing sectors where this task occurs show limited digitization of physical equipment operation, and safety-critical requirements slow automation uptake.
Sector adoption velocityclaude-sonnet-51/5Industrial/utility sectors with boiler and hopper operations are low-digitization, physical environments where AI/robotic adoption for manual equipment tasks is minimal.
Augmentation potentialclaude-haiku-4-5-202510012/5Remote monitoring dashboards and predictive maintenance alerts could assist operators by flagging blockages or wear patterns, but current AI offers limited real-time assistance during active operation or hands-on adjustment and repair work.
Augmentation potentialclaude-sonnet-52/5AI could assist with predictive maintenance alerts or diagnostic support, but it offers little direct help with the physical adjustment and repair work itself.
Task automatabilityclaude-haiku-4-5-202510012/5Operating mechanical hoppers requires physical manipulation, real-time sensing of material flow, and adaptive response to blockages or misalignment—tasks that current AI systems cannot reliably perform without specialized robotics. While remote monitoring could be partially automated, hands-on adjustment and repair remain beyond current autonomous capabilities.
Task automatabilityclaude-sonnet-51/5Operating and physically adjusting/repairing mechanical hoppers requires hands-on manipulation, physical presence, and manual dexterity that current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: operational safety regulations for boiler systems, OSHA requirements for equipment operation, and potential liability for automation failures that could damage equipment or cause facility downtime. Human operators are typically required by regulation or facility insurance.
Adoption barriersclaude-sonnet-54/5Physical equipment operation and repair often involves safety regulations, certification requirements for stationary engineers, and liability concerns that require a qualified human on-site.
Cost vs. human wageclaude-haiku-4-5-202510011/5The capital cost of robotic systems capable of hopper operation, combined with integration, setup, and ongoing maintenance, far exceeds the loaded wage of a stationary engineer performing routine hopper work.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical operation/repair task, so AI cost comparison is effectively inapplicable and human labor remains the only option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product today can autonomously operate, adjust, or repair mechanical hoppers in production environments. This task demands physical embodiment, tactile feedback, and on-site mechanical intervention that existing AI systems lack.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product operates or physically repairs mechanical hoppers; this remains a physical, hands-on task with no automation product on the market.

Clean and lubricate boilers and auxiliary equipment and make minor adjustments as needed, using hand tools.

5

CI 0–10 · exposure 0 · augmentation 25 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of automation in boiler operations remains minimal; the sector is capital-intensive, heavily regulated, and relies on skilled human operators with legal certifications who bear direct liability for safety.
Sector adoption velocityclaude-sonnet-51/5Industrial facilities maintenance and boiler operations are a low-digitization, physical-labor sector with minimal AI/robotic adoption for hands-on maintenance tasks.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by monitoring boiler telemetry and predicting maintenance windows, but the physical cleaning, lubrication, and adjustment work remains firmly in human hands, limiting the scope of augmentation.
Augmentation potentialclaude-sonnet-52/5AI could assist with predictive maintenance scheduling or diagnostic alerts, but offers little direct assistance for the physical cleaning, lubricating, and adjusting actions themselves.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires physical manipulation of hand tools on large machinery in an unstructured environment, involving tactile feedback and real-time adjustment. Current AI systems lack the embodied dexterity, environmental perception, and mechanical judgment needed to perform physical maintenance work reliably.
Task automatabilityclaude-sonnet-51/5This is a physical manual task requiring hands-on cleaning, lubrication, and hand-tool adjustments on industrial equipment, which current AI systems cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5This task carries high regulatory barriers: licensed stationary engineers and boiler operators are legally required in most jurisdictions to certify, operate, and maintain boiler systems. Liability and safety regulations mandate human accountability and sign-off.
Adoption barriersclaude-sonnet-53/5While not strictly licensed for this specific subtask, boiler operation often falls under safety regulations and certified operator oversight, and physical access/liability concerns create moderate friction against automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Robotics capable of performing this task reliably are not yet commercially available at scale; deployment costs would far exceed the wage of a trained stationary engineer doing the work manually.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for this physical maintenance task, so any hypothetical robotic solution would be far more costly than a human technician performing routine upkeep.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed commercial AI system can autonomously perform hands-on boiler maintenance with hand tools. This remains entirely dependent on human technicians with specialized training and domain expertise.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical boiler cleaning, lubrication, or hand-tool adjustments; this remains firmly in the domain of human physical labor and robotics research at best.

Receive instructions from steam engineers regarding steam plant and air compressor operations.

5

CI 0–10 · exposure 5 · augmentation 25 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Stationary engineering and boiler operations are heavily regulated, safety-critical, and physically grounded sectors with strict licensing requirements, showing minimal AI adoption and strong institutional resistance to displacement.
Sector adoption velocityclaude-sonnet-51/5Industrial/utility plant operations are a low-digitization, physical-labor sector with slow AI adoption for hands-on operational coordination tasks.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist by logging or summarizing instructions, but the core task of receiving and understanding expert guidance is inherently human-to-human communication with limited augmentation potential.
Augmentation potentialclaude-sonnet-52/5AI could help log, transcribe, or track instructions (e.g., via digital logs or voice-to-text systems), providing minor assistance, but it doesn't materially transform the core communication and response task.
Task automatabilityclaude-haiku-4-5-202510011/5This task is fundamentally about receiving and comprehending verbal or written instructions from a human expert. While AI can process text, it cannot autonomously act on instructions in a physical plant environment without human oversight, and receiving instructions is not a process that saves time when automated.
Task automatabilityclaude-sonnet-51/5Receiving verbal/written operational instructions and acting on them in a physical plant setting requires human presence, judgment, and real-time coordination that current AI cannot substitute end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5This task involves a licensed boiler operator role subject to ASME certification and state regulations requiring a qualified human to be present and responsible for steam plant operations; regulatory and safety frameworks legally mandate human judgment and accountability.
Adoption barriersclaude-sonnet-54/5Boiler and steam plant operations are subject to safety regulations and often require licensed operators, and instructions/communication in hazardous environments typically demand human accountability and legal sign-off.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying AI to manage instruction reception and comprehension would exceed the minimal wage cost of a human receiving instructions, especially when oversight and error correction are factored in.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task alone, so cost comparison favors the human by default since AI cannot deliver the output.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI systems can parse and summarize written instructions (e.g., via NLP), but no deployed product reliably receives, interprets, and acts on real-time instructions in a safety-critical industrial setting where context and domain expertise are essential.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs this interpersonal instruction-receiving and physical response task in production; it's fundamentally a workplace communication/coordination task tied to physical operations.

Provide assistance to plumbers in repairing or replacing water, sewer, or waste lines, and in daily maintenance activities.

5

CI 0–10 · exposure 0 · augmentation 25 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This is a physical, hands-on task in a traditional sector with minimal digitization. Adoption of AI for plumbing assistance is negligible because the core work cannot be performed by machines.
Sector adoption velocityclaude-sonnet-51/5Building maintenance and skilled trades are among the least digitized, lowest AI-adoption sectors, with physical labor tasks seeing negligible automation deployment.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist with diagnostics or documentation (e.g., identifying line locations via imaging or drafting maintenance records), but current systems offer limited productivity gains for the core repair and maintenance work itself.
Augmentation potentialclaude-sonnet-52/5AI could marginally help with scheduling, diagnostics documentation, or looking up part specifications, but offers little assistance to the actual physical task of aiding plumbing repairs.
Task automatabilityclaude-haiku-4-5-202510011/5This task involves hands-on physical work—assisting with repair, replacement, and maintenance of plumbing infrastructure—which requires dexterity, spatial reasoning, and real-time problem-solving in complex physical environments that current AI cannot perform. No meaningful part of end-to-end plumbing assistance can be automated today.
Task automatabilityclaude-sonnet-51/5This is hands-on physical assistance work involving manual labor, tool handling, and coordination with another tradesperson, none of which current AI systems can perform.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and safety barriers exist: plumbing work typically requires licensed plumbers or certified assistants, with liability for code violations and system failures resting on credentialed humans. Regulatory frameworks mandate human oversight of water and sewer work.
Adoption barriersclaude-sonnet-53/5While not formally licensed work itself, physical plumbing assistance often occurs in facilities with safety, code compliance, and liability considerations that favor human workers, though no strict licensing barrier applies to the helper role specifically.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot currently perform this task at all, making direct cost comparison impossible; the task remains entirely dependent on human labor with no viable AI alternative to reduce costs.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for this physical labor task, so any AI-based approach would require robotics far exceeding current cost-effectiveness compared to a human helper.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can reliably perform physical plumbing repairs or maintenance. While diagnostic AI exists for some building systems, actual repair work remains exclusively human-dependent in production environments.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product or robotic system exists that can provide physical assistance to plumbers on pipe repair or maintenance tasks in real work settings.

Supervise the work of assistant stationary engineers, turbine operators, boiler tenders, or air conditioning and refrigeration operators and mechanics.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Manufacturing, utilities, and mechanical services sectors move slowly on AI adoption; supervisory roles remain human-anchored due to safety and legal accountability requirements that prevent rapid displacement.
Sector adoption velocityclaude-sonnet-51/5Facilities and plant operations sectors are slow adopters of AI for physical supervisory roles, with minimal digitization of this specific function.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with data aggregation or alert prioritization across multiple equipment systems, but supervisory judgment—personnel decisions, safety sign-off, coaching—remains inherently human and cannot be substantially augmented by current systems.
Augmentation potentialclaude-sonnet-52/5AI could help with scheduling, monitoring dashboards, or performance tracking, but offers limited assistance to the core supervisory judgment and personnel management involved.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising technical staff requires real-time judgment, interpersonal communication, conflict resolution, and accountability for safety—inherently human responsibilities that current AI cannot discharge end-to-end. No meaningful automation pathway exists for delegating supervisory authority.
Task automatabilityclaude-sonnet-51/5Supervising skilled trade workers requires real-time physical presence, hands-on judgment, and accountability for safety-critical equipment that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Supervision of safety-critical equipment operations (boilers, turbines, refrigeration) is legally and organizationally tied to licensed, accountable human operators. Liability, worker safety regulations, and OSHA requirements create hard barriers to automation.
Adoption barriersclaude-sonnet-54/5Supervisory roles in boiler/turbine operations often require licensure, safety accountability, and legal responsibility for subordinate staff, creating strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Deploying AI systems for supervision would require extensive integration, human oversight, and liability coverage; the cost of such a system would far exceed the loaded wage of a stationary engineer supervisor.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this supervisory role, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs supervisory oversight of skilled technical teams in production environments. Monitoring logs or performance dashboards is not equivalent to the judgment, coaching, and accountability embodied in supervision.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs supervisory oversight of stationary engineers or mechanics on-site; this remains a purely human management function.

Perform or arrange for repairs, such as complete overhauls, replacement of defective valves, gaskets, or bearings, or fabrication of new parts.

3

CI 0–5 · exposure 0 · augmentation 38 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Boiler operation and repair is a heavily regulated, physically embedded, low-digital-adoption sector. Adoption of automation in this domain has been minimal because the work is inherently hands-on and safety-critical.
Sector adoption velocityclaude-sonnet-51/5Industrial maintenance and facilities engineering are low-digitization, physical-labor-heavy sectors with minimal AI/robotic adoption for hands-on repair work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist marginally by analyzing sensor data, recommending which parts to order, or helping document procedures, but the core repair execution remains entirely manual. Augmentation value is limited because the bottleneck is physical execution, not decision support.
Augmentation potentialclaude-sonnet-53/5AI can assist with diagnostics, predictive maintenance alerts, parts ordering, and repair documentation, but the physical repair itself remains unaided by AI.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires hands-on physical repair work, spatial judgment, and real-time decision-making in complex mechanical systems. Current AI cannot perform the actual fabrication, replacement, or overhaul of physical components—only a human technician or skilled worker can execute these activities.
Task automatabilityclaude-sonnet-51/5This is a physical repair and mechanical fabrication task requiring hands-on manipulation, diagnosis, and skilled trade work that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Stationary engineers and boiler operators are typically licensed professionals; regulations and safety codes mandate that repairs and overhauls be performed or directly overseen by qualified, licensed personnel. Legal liability and safety certification create hard barriers to automation.
Adoption barriersclaude-sonnet-54/5Stationary engineers/boiler operators often require licensing, and safety-critical equipment repairs carry high liability and regulatory oversight, creating strong barriers to non-human execution.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI has no role in executing the core repair task itself, so there is no meaningful cost comparison. A skilled stationary engineer's labor cost far exceeds any AI system that could only advise or document, not perform the work.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for the physical labor involved, so any AI cost comparison is moot—humans remain the only option and thus cheaper by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system can autonomously perform or arrange physical repairs, complete overhauls, or fabricate replacement parts in boiler and stationary engine systems. This remains entirely within human technician domain.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical boiler/valve repairs or fabrication; robotics for this specific unstructured industrial maintenance work remains research-stage at best.

Install burners and auxiliary equipment, using hand tools.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5The boiler operation sector remains low-digitization and heavily dependent on on-site physical work; adoption of any form of AI automation is minimal and adoption of robotic installation is virtually nonexistent.
Sector adoption velocityclaude-sonnet-51/5Industrial mechanical trades and facilities/plant operations sectors show minimal AI or robotic adoption for physical installation work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist marginally with technical documentation, safety checklists, or diagnostic guides before/after installation, but provides little support during the hands-on mechanical installation work itself.
Augmentation potentialclaude-sonnet-52/5AI could assist with diagnostics, documentation, or planning around the installation, but offers little direct help with the physical hand-tool work itself.
Task automatabilityclaude-haiku-4-5-202510011/5Installing burners and auxiliary equipment requires manual dexterity, spatial reasoning in physical environments, and real-time troubleshooting with hand tools—capabilities current AI systems lack. No end-to-end automation solution exists for this hands-on mechanical task.
Task automatabilityclaude-sonnet-51/5This is a physical hands-on installation task requiring manipulation of hardware, tools, and equipment in real-world industrial settings, which current AI cannot perform.
Adoption barriersclaude-haiku-4-5-202510015/5Boiler and burner installation typically requires certification (licensed boiler operator in many jurisdictions) and carries significant liability risk—injuries or equipment failure can be catastrophic. Legal and regulatory requirements mandate human oversight and accountability.
Adoption barriersclaude-sonnet-54/5Boiler and burner installation often requires certified/licensed technicians due to safety, code compliance, and liability concerns, creating strong regulatory and safety barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems capable of performing this physical task (if they existed) would require expensive robotic hardware and integration, far exceeding the cost of a trained human technician performing the installation.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for this physical labor, so any AI-based approach would be far more expensive or simply infeasible compared to a human technician.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs physical installation of burners and equipment. Robotics that could handle this work remain experimental and highly specialized, not available as general-purpose commercial systems.
Technical feasibility todayclaude-sonnet-51/5No deployed AI or robotic product installs burners and auxiliary equipment autonomously; this remains a manual skilled-trade task.

Ignite fuel in burners, using torches or flames.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Boiler operations and stationary engineering remain in traditional, heavily regulated, low-automation sectors where human operators are legally required and physically present on-site. Adoption of automation for critical ignition functions is negligible.
Sector adoption velocityclaude-sonnet-51/5Industrial/utility boiler operation is a slow-adopting, physically-grounded sector with low digitization of this specific hands-on task; automation here comes from control systems, not AI models.
Augmentation potentialclaude-haiku-4-5-202510012/5While AI could theoretically assist with monitoring or alerting on burner status, the actual ignition task itself offers minimal augmentation opportunity because human judgment and safety presence are legally mandated and the physical action is brief and routine.
Augmentation potentialclaude-sonnet-52/5AI could assist with monitoring, diagnostics, or predictive maintenance around boiler operations, but offers minimal direct assistance for the physical act of igniting fuel.
Task automatabilityclaude-haiku-4-5-202510011/5Igniting fuel in burners is a physical manipulation task requiring precise torch or flame control in complex environments with safety constraints. Current AI systems lack embodied robotics capabilities to reliably perform this dangerous task end-to-end in real industrial settings.
Task automatabilityclaude-sonnet-51/5This is a physical ignition procedure requiring hands-on presence at equipment; current AI cannot perform the physical act of igniting a burner, though automated ignition systems (not AI) already exist as engineering controls.
Adoption barriersclaude-haiku-4-5-202510015/5This task is heavily protected by safety regulations, industry codes, and liability requirements that typically mandate licensed human operators for burner ignition and boiler operation. Legal and insurance barriers prevent substitution with unattended automation.
Adoption barriersclaude-sonnet-54/5Boiler operation is safety-critical and often requires licensed operators under regulatory codes, and physical ignition tasks involve fire/explosion risk creating strong liability and physical-presence barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5Developing and deploying a robotic system capable of safely igniting burners would require substantial specialized hardware, safety systems, and integration costs far exceeding the wages of a human operator performing this task.
Cost vs. human wageclaude-sonnet-51/5There is no AI system that substitutes for this physical task, so no meaningful cost comparison favors AI; the applicable technology is automated hardware, not AI software.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably performs torch or flame ignition of industrial burners in production environments. This task requires physical manipulation, real-time environmental feedback, and safety compliance that current autonomous systems cannot dependably execute.
Technical feasibility todayclaude-sonnet-51/5No AI product performs physical burner ignition; this is handled by mechanical/electronic ignition systems or manual procedures, not AI-based products.

Switch from automatic to manual controls and isolate equipment mechanically and electrically to allow for safe inspection and repair work.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Industrial equipment operation is highly regulated and risk-sensitive; adoption of automation in this domain is minimal. The sector is slow to digitize and any changes face strong regulatory and organizational resistance.
Sector adoption velocityclaude-sonnet-51/5Industrial plant operations and physical equipment maintenance are low-digitization, low-automation-adoption sectors with minimal AI penetration into hands-on mechanical/electrical isolation work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist in generating or checking isolation checklists or monitoring system state, but the core tasks of physically switching controls and mechanically isolating equipment require human expertise and legal accountability that AI cannot augment meaningfully today.
Augmentation potentialclaude-sonnet-52/5AI could potentially assist with checklists, procedure documentation, or monitoring system status during the transition, but offers minimal help with the core physical isolation actions.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires physical manipulation of mechanical and electrical controls, precise isolation of equipment, and real-time assessment of system state before proceeding. Current AI systems cannot reliably perform the dexterous physical actions and safety-critical decisions involved in switching controls and isolating live equipment.
Task automatabilityclaude-sonnet-51/5This requires physical manipulation of valves, switches, and lockout/tagout equipment on-site, which current AI systems cannot perform without robotic embodiment.
Adoption barriersclaude-haiku-4-5-202510015/5Regulatory and safety codes (OSHA, NFPA, state licensing) typically require a licensed stationary engineer or qualified supervisor to personally perform and certify isolation and control switching. Liability exposure for automation of safety-critical equipment isolation is severe.
Adoption barriersclaude-sonnet-55/5Safety-critical lockout/tagout procedures are heavily regulated (OSHA) and require certified personnel physically present to isolate equipment before repair, making substitution legally and practically infeasible.
Cost vs. human wageclaude-haiku-4-5-202510011/5An AI system capable of this task would require expensive robotics, redundant safety systems, and integration costs that far exceed the loaded wage of a skilled stationary engineer performing the work.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for the physical task, so cost comparison favors the human by default since AI cannot perform the work at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI products can autonomously perform this safety-critical task involving physical equipment manipulation. This requires physical embodiment, real-time sensory feedback, and responsibility for preventing injury—capabilities that do not exist in production systems today.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical isolation and manual control switching of industrial boiler equipment; this remains firmly in the domain of human technicians.

Related occupations — Production

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.