Gas Plant Operators

51-8092.00

Median wage $87,820/yr18,030 employed (US)Rank #618 of 923 scored · top 67% by substitution

Distribute or process gas for utility companies and others by controlling compressors to maintain specified pressures on main pipelines.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution23

Exposure23

Augmentation52

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

19 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%22

panel mean rating 1.9/5 → substitution pressure 22/100

Technical feasibility todayw 20%23

panel mean rating 1.9/5 → substitution pressure 23/100

Cost vs. human wagew 15%24

panel mean rating 2.0/5 → substitution pressure 24/100

Adoption barriersw 20%inverted — strong barriers lower the score22

panel mean rating 4.1/5 (barrier strength) → substitution pressure 22/100

Sector adoption velocityw 10%21

panel mean rating 1.8/5 → substitution pressure 21/100

Task breakdown (19 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.

Test gas, chemicals, and air during processing to assess factors such as purity and moisture content, and to detect quality problems or gas or chemical leaks.

CI 25–70 · exposure 50 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Petrochemical and gas processing sectors are high-digitization industries with strong capital investment incentives; continuous automated monitoring is already standard in modern plants. Adoption is deep and accelerating in new facilities and retrofit programs.
Sector adoption velocity	claude-sonnet-5	2/5	Gas plants are capital-intensive, safety-critical industrial environments with slow technology adoption cycles relative to information-sector industries, though sensor/IoT monitoring is gradually increasing.
Augmentation potential	claude-haiku-4-5-20251001	3/5	Automated systems assist operators by continuous real-time alerting and data visualization, reducing the cognitive and manual burden of routine sampling. However, the human's role becomes mostly reactive monitoring and problem-solving rather than primary task execution, limiting transformative augmentation.
Augmentation potential	claude-sonnet-5	4/5	Continuous sensor monitoring, automated alerts, and predictive analytics significantly enhance operators' ability to detect leaks and quality issues faster and more consistently than manual testing alone.
Task automatability	claude-haiku-4-5-20251001	4/5	AI-driven sensor systems and automated analytical instruments can continuously monitor gas purity, moisture content, and detect leaks in real-time with high accuracy, achieving well over 50% time savings compared to manual testing. However, complex anomaly interpretation and response to ambiguous readings may still require human oversight, preventing a perfect 5.
Task automatability	claude-sonnet-5	2/5	Sensors and analyzers can automate data collection, but interpreting results, physically inspecting equipment, and responding to anomalies in a hazardous industrial setting still require human presence and judgment.AI cannot end-to-end replace the physical testing and on-site verification.'
Adoption barriers	claude-haiku-4-5-20251001	3/5	Regulatory frameworks (EPA, OSHA) may require documented monitoring and record-keeping, and some jurisdictions mandate human oversight of critical process safety. These create modest friction but do not categorically prevent automation—they require integration into compliance workflows rather than a human signature requirement.
Adoption barriers	claude-sonnet-5	4/5	Process safety regulations (OSHA PSM, EPA) and industry safety standards typically require qualified personnel to monitor and verify hazardous gas/chemical conditions, creating strong liability and regulatory barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	After initial capital investment in sensors and automation infrastructure, the per-task cost of automated continuous monitoring is substantially lower than staffing human operators for frequent manual testing across multiple points. Maintenance costs are moderate, yielding favorable long-run economics.
Cost vs. human wage	claude-sonnet-5	2/5	Sensor systems and monitoring software have upfront and maintenance costs comparable to or sometimes exceeding the marginal cost of trained operators already on-site for safety reasons, given required redundancy and calibration.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature automated gas detection and analysis systems (chromatography, spectroscopy, mass spectrometry) are deployed in production across refineries and chemical plants today. These systems reliably perform the core monitoring task at scale, though occasional calibration and human validation of edge cases remain common practice.
Technical feasibility today	claude-sonnet-5	2/5	Automated gas detection and monitoring sensors are deployed widely, but full replacement of the human testing/inspection task including physical sampling and judgment calls is not yet standard production practice.

Read logsheets to determine product demand and disposition, or to detect malfunctions.

CI 29–62 · exposure 45 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Oil, gas, and utility sectors have been investing in automated monitoring for decades, but adoption of modern AI-based anomaly detection (beyond traditional SCADA rules) is still in pilot and early production phases in most facilities. Legacy systems and conservative risk management slow deep, rapid rollout.
Sector adoption velocity	claude-sonnet-5	2/5	Oil and gas/utilities sectors are historically slow adopters of AI-driven monitoring compared to information/finance industries, though predictive maintenance pilots are growing.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI systems that highlight unusual log entries and flag potential malfunctions can substantially assist operators in prioritizing their attention and reducing cognitive load during long shifts. This augmentation approach (human in the loop) is already being adopted and demonstrably improves productivity and safety outcomes.
Augmentation potential	claude-sonnet-5	4/5	AI-based anomaly detection and dashboard summarization can meaningfully help operators quickly spot deviations in demand or malfunctions, improving efficiency while the operator retains final judgment.
Task automatability	claude-haiku-4-5-20251001	4/5	Log reading and pattern detection for demand and malfunction identification can be largely automated using OCR, data parsing, and anomaly detection algorithms. Current AI systems can reliably extract numerical data from logs and flag deviations from expected parameters, achieving significant time savings; however, the task may still require human judgment in ambiguous situations or for complex multi-variable interpretation.
Task automatability	claude-sonnet-5	2/5	AI can parse structured logsheet data and flag anomalies, but integrating with legacy SCADA/paper systems and interpreting context-specific operational nuances limits full end-to-end automation today.for many plants.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Gas plants operate under safety and environmental regulations (EPA, OSHA) that may require documented human oversight or sign-off on critical decisions. Liability for missed malfunctions and the expectation that operators remain alert to plant status create organizational friction, even if automated monitoring can flag issues.
Adoption barriers	claude-sonnet-5	4/5	Safety-critical operations in gas plants typically require certified operators to monitor and sign off on readings, with regulatory oversight for process safety management, creating strong human-in-the-loop requirements.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Once deployed, AI monitoring systems have very low per-incident inference costs compared to paying an operator to manually read and interpret logs continuously. Integration and maintenance costs are modest relative to the operator wage, making the cost ratio favorable.
Cost vs. human wage	claude-sonnet-5	3/5	Where digitized data feeds exist, automated monitoring software is cheaper per check than operator review, but retrofitting sensors and log digitization adds cost that narrows the gap.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Log analysis and anomaly detection systems exist in deployed monitoring software (SCADA, process control systems with alerting), but most rely on rule-based thresholds rather than AI. AI-powered anomaly detection products are available but typically require domain-specific training and tuning, limiting their plug-and-play reliability across different facilities.
Technical feasibility today	claude-sonnet-5	2/5	Some industrial monitoring/analytics products exist for anomaly detection in process data, but few are deployed specifically to replace operator log review at scale in gas plants.

Control equipment to regulate flow and pressure of gas to feedlines of boilers, furnaces, and related steam-generating or heating equipment.

CI 25–56 · exposure 42 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Gas and utility sectors have adopted automated process control extensively for decades; SCADA and DCS are industry standard. Modern deployments increasingly use AI for predictive maintenance and optimization, with adoption deeper in large utilities and refineries than smaller operations.
Sector adoption velocity	claude-sonnet-5	2/5	Energy and utilities sectors are traditionally slower adopters of cutting-edge AI compared to information/finance sectors, relying on established SCADA/DCS systems rather than novel AI agents for control tasks.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-driven monitoring dashboards, predictive alerts for equipment stress, and automated logging of anomalies significantly assist operators in managing complex systems. Augmentation is strong because AI handles continuous surveillance and pattern detection while operators focus on judgment calls and rare events.
Augmentation potential	claude-sonnet-5	3/5	AI-based predictive analytics and anomaly detection can assist operators in monitoring trends and flagging issues, improving decision-making, though the core physical control task remains human-supervised.
Task automatability	claude-haiku-4-5-20251001	3/5	Modern industrial control systems can automate flow and pressure regulation using sensors, controllers, and feedback loops—a core function of SCADA systems. However, responding to anomalies, emergency shutdowns, and equipment failures still requires human judgment and physical intervention, limiting full end-to-end automation to roughly 40–50% time savings.
Task automatability	claude-sonnet-5	2/5	This involves real-time physical control of industrial equipment with safety-critical implications; while control loops can be automated with SCADA/PLC systems, the task as described requires physical presence, sensor interpretation, and emergency response that current general AI cannot fully replace end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Gas plant operations are subject to strict safety regulations (OSHA, EPA, industry codes) that mandate operator presence, training certification, and legal responsibility for safe operation. Liability and safety-critical nature create strong regulatory and legal barriers to full autonomous operation without licensed human oversight.
Adoption barriers	claude-sonnet-5	4/5	Gas plant operations are heavily regulated for safety, often requiring certified operators, and error costs (explosions, leaks) are severe, creating strong liability and regulatory barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Industrial automation hardware and software are capital-intensive upfront but extremely cheap per operation once installed. Inference and monitoring costs are negligible compared to a loaded operator wage ($60k–$90k annually), making the long-run cost ratio heavily favorable.
Cost vs. human wage	claude-sonnet-5	2/5	Industrial control automation requires significant capital investment in specialized hardware, sensors, and safety systems, which is costly relative to a single operator's wage, though it can pay off at scale over time.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed SCADA and DCS systems reliably perform real-time flow and pressure control in production gas plants worldwide. However, they typically operate as supervisory tools requiring human operators to monitor and intervene, rather than fully autonomous end-to-end systems, so maturity is high but not purely autonomous.
Technical feasibility today	claude-sonnet-5	2/5	Automated control systems (DCS/SCADA) are deployed in gas plants for routine regulation, but full autonomous operation without human oversight is rare due to safety and reliability requirements; these are industrial control systems rather than general AI products.

Record, review, and compile operations records, test results, and gauge readings such as temperatures, pressures, concentrations, and flows.

CI 25–48 · exposure 38 · augmentation 63 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas plants are capital-intensive, legacy-heavy infrastructure with slow modernization cycles. While some automation of sensor reading collection has occurred, adoption of AI-driven review and compilation systems remains limited; most facilities still rely on human operator judgment for regulatory compliance and safety.
Sector adoption velocity	claude-sonnet-5	2/5	Gas plant operations are a heavily physical, industrial, safety-regulated sector with historically slow digitization and cautious adoption of AI-driven monitoring compared to information-sector industries.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-assisted dashboards that highlight anomalies, flag out-of-range readings, and pre-populate record templates can meaningfully assist operators in monitoring and compilation tasks. However, the safety-critical nature of the work means the human operator must retain full interpretive authority.
Augmentation potential	claude-sonnet-5	4/5	AI can significantly assist by aggregating, flagging anomalies, and summarizing large volumes of gauge and test data, improving operator efficiency while the operator retains final judgment.
Task automatability	claude-haiku-4-5-20251001	2/5	Data entry and record compilation from existing readings could be partially automated with digital sensors and logging systems, but the task requires interpreting gauge readings, reviewing operational context, and making judgment calls about anomalies—activities that demand human expertise. Current AI cannot reliably replicate the safety-critical review and decision-making components end-to-end.
Task automatability	claude-sonnet-5	3/5	Recording and compiling structured sensor/gauge data can largely be automated via SCADA/historian integration, but review requiring judgment about anomalies still needs human oversight, so only part meets the 50% bar without added AI analytics.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Gas plant operations are heavily regulated (OSHA, EPA, state environmental boards); records and gauge interpretations must be signed off by licensed operators or certified technicians. Regulatory frameworks and liability for safety-critical decisions create strong legal and professional barriers to full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement for record-keeping itself, but safety-critical process oversight in gas plants creates regulatory and liability pressure to have qualified operators verify readings.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	While sensor automation and data logging reduce some recording labor, the oversight, review, and interpretation of operations records still requires trained human operators. The all-in cost of automation (sensors, integration, AI oversight systems) often remains comparable to or exceeds the wage of a skilled operator performing these tasks.
Cost vs. human wage	claude-sonnet-5	3/5	Automated data logging is cheap, but integrating AI review of readings with plant-specific context and validation still requires engineering and oversight costs comparable to human review in many plants.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Digital sensor integration and automated logging exist in modern plants, but mature end-to-end systems that independently review, interpret, and compile records with the judgment required for safety compliance are not reliably deployed at scale. Most deployments still rely on human operators to validate and interpret readings.
Technical feasibility today	claude-sonnet-5	3/5	Industrial control systems and historians already automatically log and compile readings in production, but AI-driven review/compilation with contextual judgment is narrower and less mature in deployed products.

Monitor equipment functioning, observe temperature, level, and flow gauges, and perform regular unit checks to ensure that all equipment is operating as it should.

CI 25–39 · exposure 38 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas plants operate in heavily regulated, often conservative sectors with long asset lifecycles; while monitoring systems are being modernized, widespread AI-driven autonomous monitoring remains limited, with most facilities still relying on traditional operator rounds and manual record-keeping.
Sector adoption velocity	claude-sonnet-5	2/5	Oil and gas is a moderately slow-adopting sector for AI in physical operations, with automation focused on data collection and alerts rather than full task replacement.
Augmentation potential	claude-haiku-4-5-20251001	4/5	Real-time alerting systems, predictive analytics on sensor data, and automated anomaly detection significantly enhance operator situational awareness and reduce manual gauge-checking burden, enabling faster response to problems while the human operator remains responsible for judgment and action.
Augmentation potential	claude-sonnet-5	4/5	AI-enabled sensor analytics, predictive maintenance, and anomaly detection significantly help operators catch issues faster and prioritize checks, meaningfully boosting productivity while humans remain responsible for physical verification.
Task automatability	claude-haiku-4-5-20251001	3/5	Monitoring gauges and checking equipment status can be partially automated via sensor networks and anomaly detection systems, but real-time physical inspections and contextual judgment about emerging faults still require human intervention, limiting full end-to-end automation to roughly 50% time savings.
Task automatability	claude-sonnet-5	2/5	Sensor monitoring can be augmented with automated alarms and analytics, but the physical presence, integrated judgment, and unit walk-throughs required for gas plant operations still require significant human involvement today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety-critical infrastructure regulations (EPA, OSHA, industry standards) mandate operator presence and qualified human sign-off on equipment status; liability for failures and the legal requirement for licensed operators to monitor hazardous processes creates strong regulatory barriers to full substitution.
Adoption barriers	claude-sonnet-5	4/5	Gas plants are heavily regulated, requiring certified operators for safety-critical monitoring, with liability and regulatory frameworks mandating human oversight and unit inspections.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Industrial automation systems (sensors, SCADA, cloud integration) require significant upfront capital and ongoing maintenance costs, and operator oversight remains necessary, making the all-in cost comparable to or exceeding a single operator's loaded wage in many plants.
Cost vs. human wage	claude-sonnet-5	2/5	Sensor and analytics systems have upfront and maintenance costs comparable to or exceeding incremental labor savings, especially since human presence is still needed for physical checks and safety compliance.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Industrial monitoring systems and IoT sensor platforms exist and are deployed in some plants, but they typically require human oversight for alarm interpretation and maintenance decisions; no fully autonomous system reliably handles the full scope of observation and decision-making at production scale.
Technical feasibility today	claude-sonnet-5	2/5	SCADA/DCS systems and predictive maintenance software are deployed in industrial plants, but full autonomous monitoring replacing operator rounds and judgment is not yet standard practice.

Contact maintenance crews when necessary.

CI 25–34 · exposure 25 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas and utilities are moderate digitizers with strong conservative/safety cultures. Alerting systems are common, but autonomous crew-contact decisions remain rare in production; most facilities require operator initiation and approval.
Sector adoption velocity	claude-sonnet-5	2/5	Gas plant operations are a heavy-industrial, safety-critical sector with historically slow AI adoption relative to information/professional services sectors.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist by analyzing sensor logs, flagging anomalies, and suggesting when to contact crews, helping operators prioritize and draft messages. This augmentation is valuable but does not transform the core task, which remains human-initiated and human-decided.
Augmentation potential	claude-sonnet-5	3/5	Automated monitoring and alert systems can flag anomalies and suggest maintenance needs, helping operators decide faster, though the actual contact and coordination remains human-driven.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can draft communications and identify when maintenance is needed based on sensor data, but the decision to contact crews—which depends on judgment about urgency, priority, and operational context—and the actual contact initiation require human oversight. Perhaps 20–30% of the workflow could be automated (flagging alerts, drafting messages), but not the full task.
Task automatability	claude-sonnet-5	2/5	Deciding when and whom to contact requires situational judgment tied to plant conditions; AI could draft or route messages but the core decision-making and escalation judgment resists full automation today.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Gas plant operations are heavily regulated; maintenance decisions and crew dispatch often require documented authorization by licensed operators and may be subject to safety protocols that mandate human accountability. Legal and safety liability fall on the operator signing off, not on an automated system.
Adoption barriers	claude-sonnet-5	3/5	Safety-critical plant operations often require human sign-off and operator accountability for escalation decisions, though this isn't a hard licensing requirement specifically for this narrow task.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI system integration, monitoring infrastructure, and human oversight (a skilled operator must still validate and authorize contact decisions) approach or exceed the cost of having the operator make the contact directly. No clear cost advantage emerges.
Cost vs. human wage	claude-sonnet-5	3/5	Automated alert systems are cheap to run, but the human judgment and accountability layer around initiating maintenance contact still requires operator time, keeping costs roughly comparable rather than dramatically lower.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably makes autonomous decisions to contact maintenance crews in gas plants. Alert systems and monitoring exist, but they still route to human operators who decide whether and when to contact crews. Automation requires legal and safety sign-off that hasn't matured.
Technical feasibility today	claude-sonnet-5	2/5	Notification/alerting software and ticketing systems exist and are used in industrial settings, but autonomous decision-making about when maintenance intervention is warranted is not a mature deployed product function.

Control operation of compressors, scrubbers, evaporators, and refrigeration equipment to liquefy, compress, or regasify natural gas.

CI 25–25 · exposure 25 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Despite digitalization trends, gas plant operations remain in a highly regulated, safety-critical domain where automation adoption is cautious and incremental. Most plants are decades old, and operators are entrenched; pilot projects exist but production displacement is minimal.
Sector adoption velocity	claude-sonnet-5	2/5	Oil and gas processing is a capital-intensive, safety-regulated sector with slower AI adoption compared to information/professional services, though some plants use advanced process control and predictive maintenance tools.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-driven dashboards, real-time anomaly alerts, and predictive maintenance recommendations can meaningfully assist operators in monitoring equipment health and optimizing settings, but the human operator remains essential for judgment and manual control.
Augmentation potential	claude-sonnet-5	3/5	AI-based predictive maintenance, anomaly detection, and process optimization tools can meaningfully assist operators in monitoring and fine-tuning equipment performance, improving efficiency while humans retain control authority.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can monitor sensor data and optimize some parameters, the task requires real-time decision-making under variable conditions, manual equipment adjustments, and emergency response that current systems cannot fully handle autonomously. The heterogeneous equipment landscape and need for physical intervention place this well below the 50% time-saving threshold.
Task automatability	claude-sonnet-5	2/5	This is a hands-on physical control task involving real-time equipment adjustment with safety-critical implications; current AI can assist monitoring but cannot autonomously operate the physical control loops end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety regulations, EPA compliance, process licensing, and insurance liability require that licensed operators maintain direct authority over critical equipment. Legal and regulatory frameworks mandate human sign-off on hazardous material handling and emergency response.
Adoption barriers	claude-sonnet-5	4/5	Gas plant operations are subject to strict safety regulations, licensing, and liability requirements that generally mandate qualified human operators to be in the loop or on-site for control and emergency response.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration of specialized sensors, control software, redundant safety systems, and continuous oversight costs are substantial. The requirement for safety-critical validation and the need for human backup operators make the all-in cost comparable to or exceeding that of a skilled operator.
Cost vs. human wage	claude-sonnet-5	2/5	Industrial control systems and sensors require significant capital investment, specialized integration, and continuous human oversight, making all-in AI costs comparable to or higher than retaining trained operators for this safety-critical role.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some industrial control systems use AI-assisted monitoring and anomaly detection, but no deployed product reliably automates the full operation of compressor/scrubber/refrigeration systems end-to-end. Production deployments exist only for narrow subcomponents (predictive maintenance alerts), not the complete operational task.
Technical feasibility today	claude-sonnet-5	2/5	Advanced process control and predictive analytics exist in industrial settings, but fully autonomous control of liquefaction/compression trains without human operators is not deployed at scale due to safety and reliability requirements.

Determine causes of abnormal pressure variances, and make corrective recommendations, such as installation of pipes to relieve overloading.

CI 25–25 · exposure 25 · augmentation 63 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas plants are capital-intensive, operationally mature infrastructure with conservative modernization. While remote monitoring and data analytics are increasingly deployed, autonomous decision-making on corrective actions remains limited; adoption skews toward augmentation and alerts rather than AI-driven recommendations.
Sector adoption velocity	claude-sonnet-5	2/5	Industrial gas/process plants are a lower-digitization, physically-oriented sector where AI adoption for core operational decision-making is still in early pilot stages compared to information-sector industries.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI excels at real-time pressure trend analysis, anomaly flagging, and pattern matching across historical data, which significantly assists operators in identifying suspect equipment and narrowing diagnostic scope. This augmentation substantially improves decision speed and coverage without replacing expert judgment on root cause and remediation design.
Augmentation potential	claude-sonnet-5	3/5	AI-based anomaly detection and predictive analytics can help operators spot pressure irregularities faster and suggest possible causes, meaningfully aiding diagnosis even though the operator must verify and decide on corrective action.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can analyze pressure data and identify some anomalies, determining root causes of abnormal variances requires integration of real-time sensor data, equipment history, site-specific configurations, and safety-critical judgment. Current systems cannot reliably perform the full diagnostic and recommendation loop autonomously—human expertise in interpreting complex failure modes and validating recommendations is essential.
Task automatability	claude-sonnet-5	2/5	Diagnosing abnormal pressure variances requires integrating live sensor data, physical plant knowledge, and equipment-specific judgment that current AI cannot fully replicate end-to-end without heavy customization and human validation.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Gas plant operations are heavily regulated (EPA, OSHA, state environmental agencies) and safety-critical; pressure management directly affects public safety. Liability and regulatory frameworks typically require licensed professional engineers to design and sign off on corrective measures like pipe installation, creating a hard barrier to full automation.
Adoption barriers	claude-sonnet-5	4/5	Gas plant operations are heavily regulated with safety-critical requirements, often mandating licensed/certified personnel for corrective actions and sign-off on structural changes like pipe installation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI-driven monitoring and alerting can reduce costs, but expert gas plant operators command high wages and the downstream engineering work (design, installation validation) requires licensed professionals. The all-in cost of AI diagnosis plus required human review and engineering sign-off remains comparable to or higher than direct human expert assessment.
Cost vs. human wage	claude-sonnet-5	2/5	Deploying and maintaining specialized industrial AI monitoring systems with sensor integration, model tuning, and safety oversight is costly relative to a skilled operator's incremental diagnostic effort, though it can scale over time.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Pressure monitoring systems with anomaly detection exist in production, but autonomous root-cause diagnosis and engineering recommendations for physical interventions remain largely at the advisory/augmentation stage. No deployed AI product reliably performs end-to-end causal diagnosis and physical remediation design in gas plant operations today.
Technical feasibility today	claude-sonnet-5	2/5	Some anomaly-detection and predictive-maintenance products exist for industrial process monitoring, but reliable autonomous root-cause diagnosis and engineering recommendations for gas plants remain limited to decision-support tools, not full replacements.

Calculate gas ratios to detect deviations from specifications, using testing apparatus.

CI 25–25 · exposure 25 · augmentation 63 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas plant operations remain moderately digitized with slow adoption of autonomous systems in production. Most facilities still rely on human operators with computer-assisted tools rather than fully automated or agent-based approaches, reflecting the safety-critical and regulated nature of the sector.
Sector adoption velocity	claude-sonnet-5	2/5	Industrial gas processing is a physical, safety-critical sector with historically slow digitization and AI adoption compared to information-based industries.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI tools can meaningfully assist operators by automating ratio calculations, alerting to deviations, and visualizing trends from test data, raising their efficiency and decision-speed. However, the human operator remains essential for apparatus operation, contextual judgment, and regulatory sign-off.
Augmentation potential	claude-sonnet-5	4/5	AI-based analytics and monitoring dashboards can significantly assist operators by flagging anomalies and suggesting likely causes, improving speed and accuracy of deviation detection while the operator retains control.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can assist with ratio calculations and pattern detection from test data, but cannot independently operate physical testing apparatus or interpret instrument readings in real-world plant conditions without human supervision. The task requires hands-on operation of specialized equipment and contextual judgment that current AI systems cannot fully replace.
Task automatability	claude-sonnet-5	2/5	The calculation portion could be automated, but the task requires physical operation of testing apparatus and integration with real-time plant sensor data, which current general AI systems cannot perform end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Regulatory frameworks (EPA, OSHA, pipeline safety regulations) typically require licensed or certified operators to perform gas quality testing and sign off on compliance. Liability for incorrect gas ratio measurements creates high error-cost asymmetry, and safety-critical applications generally require human accountability and sign-off.
Adoption barriers	claude-sonnet-5	4/5	Gas plant operations are heavily regulated for safety, often requiring certified operators to interpret readings and take corrective action, creating strong liability and regulatory barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration of AI systems for this task would require significant setup, sensor integration, and validation costs. The loaded wage of a skilled gas plant operator is substantial, and current AI solutions are not yet order-of-magnitude cheaper when accounting for apparatus interfacing, safety compliance, and oversight requirements.
Cost vs. human wage	claude-sonnet-5	2/5	Specialized industrial sensor integration, calibration, and safety-critical monitoring systems have high upfront and maintenance costs that often exceed the marginal cost of a trained operator performing routine checks.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI can perform calculations on structured data inputs, deployed products do not reliably perform the full task of autonomously operating testing apparatus and detecting deviations in production gas plant environments. Some software tools exist for data analysis but do not handle the physical apparatus operation or real-time decision-making.
Technical feasibility today	claude-sonnet-5	2/5	While SCADA/DCS systems and statistical process control software can flag deviations, fully autonomous AI-driven gas ratio calculation and interpretation in production plants remains limited and requires human verification.

Adjust temperature, pressure, vacuum, level, flow rate, or transfer of gas to maintain processes at required levels or to correct problems.

CI 20–25 · exposure 25 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas plant operations remain highly regulated and risk-averse sectors where automation has been limited to monitoring and advisory functions rather than autonomous control. Large capital investments and safety culture favor incremental adoption of decision-support tools over wholesale replacement of human operators.
Sector adoption velocity	claude-sonnet-5	2/5	Oil and gas processing is a capital-intensive, safety-critical sector with slower digitization and AI adoption compared to information/professional services, though some large facilities pilot advanced automation.
Augmentation potential	claude-haiku-4-5-20251001	4/5	Real-time AI-driven dashboards, predictive alerts for temperature/pressure anomalies, and automated data logging significantly augment operator situational awareness and reduce manual calculation and record-keeping. Modern control systems help operators respond faster and more consistently to process deviations while maintaining oversight.
Augmentation potential	claude-sonnet-5	4/5	AI-driven predictive analytics, anomaly detection, and process optimization tools can significantly assist operators in monitoring trends and flagging issues before they become critical, improving decision-making while humans remain responsible for final adjustments.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can monitor sensor data and recommend adjustments, safely executing control changes in hazardous gas environments requires real-time sensor integration, fail-safe override capability, and handling of edge cases that current autonomous systems cannot reliably manage end-to-end without human oversight. Modern SCADA systems provide decision support but operators retain direct control.
Task automatability	claude-sonnet-5	2/5	This is a real-time physical control task requiring sensor interpretation and valve/equipment adjustment in a hazardous industrial environment; current AI can assist via control algorithms but cannot end-to-end perform the physical adjustments and emergency judgment reliably today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Federal and state regulations (OSHA, EPA, state utility commissions) legally mandate that licensed, certified operators conduct and approve critical adjustments in gas processing facilities. Liability for equipment damage, explosions, or safety incidents creates hard barriers to full automation regardless of technical capability.
Adoption barriers	claude-sonnet-5	4/5	Gas plant operations are subject to strict safety regulations, licensing requirements, and liability concerns; regulatory bodies typically mandate qualified human operators to be responsible for critical process safety decisions.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	SCADA and process control automation systems are capital-intensive and require integration, training, and ongoing maintenance. When amortized across shift operations, AI-assisted monitoring can reduce labor time but does not yet achieve cost parity with skilled operator wages when accounting for certification, liability insurance, and system redundancy requirements.
Cost vs. human wage	claude-sonnet-5	2/5	Implementing advanced control systems, sensors, and safety interlocks requires significant capital investment, integration, and ongoing calibration, making the all-in cost comparable to or higher than retaining human operators for many facilities.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Deployed industrial control systems can automate routine setpoint adjustments and alert operators to anomalies, but they operate under strict human-in-the-loop constraints due to safety criticality and liability exposure. No current product independently controls gas plant operations without certified operator authorization and override.
Technical feasibility today	claude-sonnet-5	2/5	Advanced process control (APC) and SCADA-integrated optimization systems exist and are deployed in some plants, but full autonomous adjustment without human oversight is not standard practice; most facilities still require operator intervention for corrections.

Control fractioning columns, compressors, purifying towers, heat exchangers, and related equipment to extract nitrogen and oxygen from air.

CI 14–25 · exposure 17 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Gas plants are capital-intensive, long-lifecycle facilities with conservative operational cultures and stringent regulatory oversight. Adoption of autonomous or AI-driven control remains nascent; most sites still rely on traditional DCS with human operators.
Sector adoption velocity	claude-sonnet-5	2/5	Heavy industrial/manufacturing sectors like gas processing adopt automation more slowly than digital-native industries, with existing SCADA/DCS systems representing incremental rather than AI-driven transformation.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can augment operator capability through real-time monitoring dashboards, predictive alerts for equipment stress, and optimization recommendations, which improve situational awareness and response time. However, augmentation is limited to decision support while humans retain direct physical control.
Augmentation potential	claude-sonnet-5	3/5	AI-based predictive maintenance, anomaly detection, and process optimization tools can meaningfully assist operators in monitoring equipment performance and predicting failures, improving efficiency while humans remain in control.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can monitor sensor data and suggest parameter adjustments, the task requires real-time physical equipment control in a safety-critical environment with complex, non-linear interactions between columns, compressors, and heat exchangers. Current AI lacks reliable autonomous control of such interdependent physical systems without human oversight.
Task automatability	claude-sonnet-5	1/5	This involves real-time physical control of industrial equipment (fractioning columns, compressors, heat exchangers) requiring sensor monitoring and manual/valve adjustments in a physical plant environment that current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Gas plant operations face strong regulatory barriers (EPA, OSHA, state safety codes), insurance requirements, and liability asymmetry—equipment failure or safety incidents can cause injury or environmental damage. Plant safety protocols typically require licensed human operators to maintain direct control and accountability.
Adoption barriers	claude-sonnet-5	4/5	Safety-critical industrial gas processing involves regulatory oversight (OSHA, EPA), liability for equipment failures/explosions, and typically requires certified operators to monitor and intervene, creating strong barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration of AI control systems for industrial gas plants requires substantial capital investment, specialized hardware, validation, and ongoing maintenance. This exceeds the loaded wage of a single operator in most scenarios.
Cost vs. human wage	claude-sonnet-5	2/5	While automated control systems can reduce labor needs somewhat, they require significant capital investment, specialized engineering, and ongoing human oversight, making the all-in cost comparable to or only modestly better than human operators.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some monitoring and alerting systems exist, but no deployed product reliably performs autonomous control of fractioning columns and compressor networks at production scale. AI can assist with diagnostics, but the equipment operation itself remains human-controlled in practice.
Technical feasibility today	claude-sonnet-5	2/5	Advanced process control (APC) and DCS automation systems exist and are deployed in some air separation plants, but full autonomous control without human operators overseeing safety-critical equipment is not standard practice.

Monitor transportation and storage of flammable and other potentially dangerous products to ensure that safety guidelines are followed.

CI 9–25 · exposure 17 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	The chemicals and petrochemical sector is highly regulated with conservative safety cultures; automation of operator judgment has seen minimal adoption due to liability concerns, regulatory mandates for human responsibility, and the catastrophic-failure cost structure that deters substitution.
Sector adoption velocity	claude-sonnet-5	2/5	Gas and chemical plant operations are capital-intensive, physically-oriented industrial sectors with historically slower AI adoption compared to information/professional services, though sensor-based monitoring is gradually increasing.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist by highlighting anomalies, predicting maintenance needs via sensor fusion, and generating alerts faster than manual scanning, meaningfully raising operator productivity in pattern detection and record-keeping while the human retains safety decision authority.
Augmentation potential	claude-sonnet-5	4/5	AI-enabled sensor analytics, predictive maintenance, and anomaly detection significantly enhance operators' ability to monitor for safety issues and flag potential hazards in real time.
Task automatability	claude-haiku-4-5-20251001	2/5	Some aspects like sensor monitoring and data logging can be automated, but the task critically requires human judgment in real-time risk assessment, emergency response decisions, and interpretation of unusual conditions that deviate from normal parameters—elements where current AI systems lack the situational reasoning and accountability needed for chemical safety.
Task automatability	claude-sonnet-5	1/5	This requires physical presence, sensor monitoring integration, and real-time judgment in hazardous physical environments that current AI cannot perform end-to-end without extensive human oversight and physical action.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong regulatory barriers exist: EPA, OSHA, and DOT regulations typically require licensed/certified human operators to be physically present and responsible for safety monitoring of hazardous materials; liability for failures rests on the responsible human, creating a hard legal requirement for human sign-off.
Adoption barriers	claude-sonnet-5	5/5	Safety-critical monitoring of flammable/hazardous materials is heavily regulated by OSHA, EPA, and industry-specific safety codes typically requiring certified human operators to be accountable for compliance and emergency response.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Monitoring systems and sensors are expensive to install and maintain, and the cost of oversight plus AI infrastructure approaches or exceeds the loaded wage of a single operator, especially when factoring in liability and redundancy requirements for safety-critical monitoring.
Cost vs. human wage	claude-sonnet-5	2/5	Sensor and monitoring infrastructure plus AI analytics require significant capital investment and ongoing human oversight, making all-in costs comparable to or higher than human operator costs when accounting for liability and integration.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Industrial monitoring systems exist that track sensors and alert to anomalies, but no deployed AI system reliably performs the full task of safety monitoring and guideline enforcement independently; human operators remain required for judgment calls, emergency decisions, and regulatory compliance verification in production plants.
Technical feasibility today	claude-sonnet-5	2/5	Deployed SCADA and sensor-based monitoring systems exist and assist with anomaly detection, but no product autonomously monitors and ensures compliance with safety guidelines for hazardous material transport/storage without human operators.

Start and shut down plant equipment.

CI 9–20 · exposure 20 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Gas plant operations are in a heavily regulated, safety-critical sector with minimal AI agent adoption in production. Operators are unionized in many regions and cultural/organizational barriers to full automation are substantial, with adoption limited to monitoring assistance rather than autonomous control.
Sector adoption velocity	claude-sonnet-5	2/5	Energy and utilities sectors adopt automation steadily but cautiously due to safety-critical nature and regulatory scrutiny, with full autonomous control still uncommon relative to fast-moving digital sectors.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Current AI tools can assist with monitoring dashboards and alert summarization, but the task itself—physically initiating and halting equipment through control interfaces—offers limited scope for augmentation. AI does not meaningfully enhance an operator's core capability to safely start and shut down equipment.
Augmentation potential	claude-sonnet-5	3/5	AI-driven predictive analytics, anomaly detection, and automated sequencing support operators in monitoring and executing start/shutdown procedures more efficiently, though humans remain essential for judgment and safety oversight.
Task automatability	claude-haiku-4-5-20251001	2/5	Starting and shutting down plant equipment involves multiple safety interlocks, manual verification steps, and real-time operator judgment about plant state. While some monitoring and initiation sequences could be partially automated, the full task requires human judgment for anomaly detection and safety confirmation that AI cannot reliably execute end-to-end today.
Task automatability	claude-sonnet-5	2/5	Physical equipment start-up and shutdown requires on-site manipulation of valves, switches, and safety checks that current AI cannot perform end-to-end without embodied robotics; automation here is largely control-system logic, not full task substitution.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong regulatory and legal barriers exist: operators must hold certifications (e.g., Gas Plant Operator licenses), equipment startup/shutdown is covered under safety and environmental regulations, and liability for equipment damage or safety events falls on the facility. Human sign-off is typically required by OSHA and state/federal codes.
Adoption barriers	claude-sonnet-5	5/5	Gas plant operations are heavily regulated for safety (OSHA, PHMSA, etc.), often requiring licensed/certified operators to be present and responsible for equipment start-up and shutdown, with severe liability for errors.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Automation would require significant specialized hardware integration, cybersecurity hardening, and ongoing safety validation that far exceeds the cost of an experienced operator. The liability and integration costs make AI substantially more expensive than human operator labor for this mission-critical function.
Cost vs. human wage	claude-sonnet-5	2/5	Industrial control automation exists but requires significant capital investment in sensors, actuators, and safety-certified systems, plus ongoing human oversight, making all-in cost savings modest compared to a human operator's wage.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs autonomous startup/shutdown of industrial gas plant equipment in production. This task requires integration with proprietary control systems, real-time sensor validation, and safety certification—all deployed systems still require human operators to execute and sign off on these critical procedures.
Technical feasibility today	claude-sonnet-5	2/5	SCADA/DCS systems and programmable logic controllers already automate many sequenced start/stop procedures, but human operators remain in the loop for verification, exception handling, and physical safety checks in deployed plants.

Distribute or process gas for utility companies or industrial plants, using panel boards, control boards, and semi-automatic equipment.

CI 0–25 · exposure 13 · augmentation 50 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Utility and industrial gas sectors have traditionally been slow to adopt fully autonomous AI systems due to safety, regulatory, and liability constraints. While incremental monitoring and alert systems are deployed, deep operational displacement remains rare and hesitant in practice.
Sector adoption velocity	claude-sonnet-5	1/5	Utility and industrial plant operations are a slow-adopting, highly regulated, physical-infrastructure sector with minimal AI agent deployment for direct control tasks.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-assisted monitoring (anomaly detection, trend analysis, predictive alerts on panel data) can meaningfully help operators prioritize attention and accelerate routine decision-making, though the human operator must retain oversight and final control authority for safety-critical actions.
Augmentation potential	claude-sonnet-5	3/5	AI-driven analytics and predictive maintenance tools can assist operators in monitoring trends and anomaly detection on control board data, improving situational awareness without replacing hands-on control.
Task automatability	claude-haiku-4-5-20251001	2/5	While monitoring and some control decisions could be partially automated, the task requires real-time decision-making in response to variable plant conditions, safety-critical judgments, and equipment anomalies that current AI cannot reliably handle end-to-end without human oversight. Semi-automatic equipment already exists but human operators remain essential for fault diagnosis and emergency response.
Task automatability	claude-sonnet-5	1/5	This is a physical process-control task requiring real-time monitoring of industrial equipment and manual/semi-automatic adjustments; current AI cannot physically operate valves, panels, or respond to on-site emergencies.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong regulatory and safety barriers protect this task: utility commissions and industrial safety codes often mandate qualified, licensed operators with on-site presence; liability for gas-system failures (explosions, gas leaks, service interruptions) falls directly on operators and plants, making algorithmic control legally and financially risky. Human sign-off is typically required for critical distribution decisions.
Adoption barriers	claude-sonnet-5	5/5	Gas utility operations are heavily regulated, requiring certified operators, safety compliance, and liability accountability for hazardous processes, making licensed human oversight legally mandatory.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration of advanced automation into industrial plants is capital-intensive; ongoing AI system maintenance, validation, and safety certification add substantial cost. Current human operator wages in utility/industrial sectors are modest relative to the infrastructure investment required for full autonomous gas plant operation.
Cost vs. human wage	claude-sonnet-5	1/5	Replacing a licensed operator would require extensive safety-certified automation infrastructure and redundant fail-safes, making full AI substitution far more costly than current human labor.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Although SCADA systems and industrial automation software exist in deployed settings, they perform monitoring and routine control—not the full scope of gas distribution decision-making. No end-to-end AI system reliably manages the complex, variable conditions and safety-critical exceptions this task entails without significant human intervention and sign-off.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product autonomously operates gas distribution panel boards end-to-end; existing SCADA/automation systems are rule-based control systems, not AI agents performing the operator role.

Collaborate with other operators to solve unit problems.

CI 7–16 · exposure 5 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Gas plants are legacy infrastructure with slow modernization cycles and strong safety cultures that favor proven human expertise over automation of safety-critical collaboration. Adoption of AI agents in these control rooms remains minimal despite digitization elsewhere.
Sector adoption velocity	claude-sonnet-5	2/5	Process manufacturing and energy plants are relatively slow adopters of AI for operational tasks, with automation focused on monitoring/instrumentation rather than interpersonal collaboration.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist operators by flagging anomalies, suggesting diagnostics, or surfacing historical solutions to similar problems, improving problem-solving speed. However, the core collaborative judgment and accountability remain human-centered, limiting transformative augmentation potential.
Augmentation potential	claude-sonnet-5	3/5	AI-based monitoring, diagnostics, and knowledge-base tools can support operators by surfacing data and suggesting causes, aiding but not replacing their collaborative problem-solving discussions.
Task automatability	claude-haiku-4-5-20251001	1/5	Solving unit problems collaboratively requires real-time judgment about physical plant states, negotiation of competing operational priorities, and contextual decision-making among human operators. Current AI cannot dynamically participate in team problem-solving or take autonomous responsibility for physical systems.
Task automatability	claude-sonnet-5	1/5	This requires real-time human collaboration, situational awareness of physical plant conditions, and joint judgment among operators on-site; AI cannot substitute for the collaborative human problem-solving process itself.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Gas plant operation is heavily regulated (FERC, EPA, state utilities commissions), and operators must be licensed professionals with legal accountability for safe plant operation. Collaboration between operators is a safety-critical function, and regulatory frameworks require human professional judgment and sign-off.
Adoption barriers	claude-sonnet-5	4/5	Safety-critical industrial operations require certified, trained personnel physically present, with regulatory and liability frameworks discouraging removal of human oversight from collaborative troubleshooting.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI diagnostic tools are relatively cheap, but the human operator labor for collaboration is specialized and irreplaceable; AI would require significant integration and oversight costs without eliminating the human operator role, making the cost ratio unfavorable for full substitution.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI system replacing this task, so cost comparison favors the human-only baseline; any AI tool would be additive cost, not a substitute.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI can assist with diagnostic suggestions or data analysis, but no deployed system reliably collaborates as an operator peer in a control room solving real unit problems. Products exist for narrowly scoped diagnostics, but not for genuine operator-level collaboration.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs collaborative unit troubleshooting between operators; at most AI provides decision-support data, not the collaboration task itself.

Change charts in recording meters.

CI 0–24 · exposure 8 · augmentation 0 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Gas plant operations remain physically embedded and traditional; adoption of industrial robots in this sector is slow and limited to large facilities with ongoing investment in automation. Most routine tasks, including chart changes, remain manual.
Sector adoption velocity	claude-sonnet-5	1/5	Gas plant operations are a low-digitization, physical industrial sector with slow AI/robotics adoption for routine manual upkeep tasks like this.
Augmentation potential	claude-haiku-4-5-20251001	1/5	AI offers minimal assistance for a straightforward, hands-on physical task that requires no analysis, decision-making, or information processing prior to execution.
Augmentation potential	claude-sonnet-5	1/5	AI offers essentially no assistance for the physical act of changing paper charts in a recording meter.
Task automatability	claude-haiku-4-5-20251001	1/5	This is a physical task requiring hands-on manipulation of mechanical equipment in an industrial setting. Current AI systems cannot physically access, locate, and swap charts in recording meters without specialized robotic hardware that is not widely deployed in gas plants.
Task automatability	claude-sonnet-5	2/5	This is a simple physical task of swapping paper charts in a mechanical recorder, requiring on-site manual manipulation that current AI cannot perform end-to-end without robotics.; only minor automation via digital sensor replacement is feasible but not the task as stated.:
Adoption barriers	claude-haiku-4-5-20251001	5/5	Gas plant operations are heavily regulated (e.g., OSHA, EPA, industry-specific safety standards), and maintenance tasks directly affect operational safety and regulatory compliance. Legal and safety requirements strongly mandate qualified human operators perform or directly supervise such maintenance work.
Adoption barriers	claude-sonnet-5	2/5	No licensing barrier specifically prevents automation, but physical plant access, safety protocols, and equipment-specific procedures create moderate organizational friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of deploying robotics capable of performing this task safely in a gas plant far exceeds the labor cost of a technician performing it manually. Integration, safety validation, and oversight for automated systems would be prohibitively expensive.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI-based substitute for physically swapping meter charts, so any AI cost comparison is not applicable and the human remains the only viable cost option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI products perform this physical maintenance task reliably in production. This requires dexterous robotic systems operating in industrial environments, which remain research-stage or highly specialized rather than off-the-shelf solutions.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product physically changes recording meter charts; this remains a manual maintenance task in industrial plants.

Clean, maintain, and repair equipment, using hand tools, or request that repair and maintenance work be performed.

CI 0–5 · exposure 0 · augmentation 25 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Gas plant operations remain capital-intensive, safety-critical, and slow to adopt unproven automation. The sector shows laggard patterns: small specialized teams, extensive regulatory scrutiny, and heavy reliance on experienced human operators for hands-on work.
Sector adoption velocity	claude-sonnet-5	1/5	Industrial gas plant operations are a physical, safety-critical, low-digitization sector where AI adoption for hands-on maintenance tasks is minimal to nonexistent.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI tools could provide diagnostic support or maintenance scheduling recommendations, but the hands-on nature of cleaning, maintaining, and repairing equipment limits meaningful augmentation. Most value would come from assisted planning rather than real-time task assistance.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with diagnostics, predictive maintenance scheduling, or documenting issues, but offers little direct help with the physical act of cleaning and repairing equipment using hand tools.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical manipulation of hand tools and direct equipment contact in a gas plant environment. Current AI systems cannot perform hands-on repair and maintenance work, and no robot-integrated solution is widely deployed for general gas plant equipment maintenance at scale.
Task automatability	claude-sonnet-5	1/5	This is physical manual labor (cleaning, tool use, hands-on repair) that current AI systems cannot perform; no robotic system can substitute for this hands-on maintenance work today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Gas plant operations are heavily regulated; equipment repair and maintenance often require licensed technicians and regulatory compliance documentation. Safety-critical work in hazardous environments creates strong liability barriers and legal requirements for human oversight and sign-off.
Adoption barriers	claude-sonnet-5	4/5	Gas plant equipment maintenance is subject to safety regulations, certification requirements, and liability concerns that typically require qualified human operators/technicians to perform or sign off on repairs.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The physical nature of this work, combined with the specialized environment and safety requirements of gas plants, makes current automation prohibitively expensive relative to skilled operator wages. Custom robotics would far exceed the cost of human labor for this task.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute for physical tool-based repair work, so AI cost is effectively infinite relative to the human doing the manual task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product performs end-to-end cleaning, maintenance, and repair of gas plant equipment. Specialized industrial robots exist for narrow, repetitive tasks, but general-purpose maintenance with hand tools in complex industrial settings remains beyond current capability.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical cleaning/repair of gas plant equipment; this remains firmly in the domain of human technicians.

Signal or direct workers who tend auxiliary equipment.

CI 0–5 · exposure 0 · augmentation 25 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Gas plant operations remain heavily regulated and safety-critical; adoption of AI for worker direction is minimal. The sector is conservative on automating safety-related human interactions.
Sector adoption velocity	claude-sonnet-5	1/5	Gas plant operations are a heavy-industry, low-digitization sector where AI adoption for physical coordination and supervisory signaling tasks is minimal and not seeing measurable production deployment.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist by logging or analyzing equipment status to inform what signals an operator should give, but the core act of directing workers in real time remains the operator's responsibility, limiting augmentation scope.
Augmentation potential	claude-sonnet-5	2/5	AI-enabled monitoring systems and communication tools (e.g., automated alerts, sensor dashboards) can support situational awareness for the operator, but they don't substantially transform the direct interpersonal signaling and directing of auxiliary equipment workers.
Task automatability	claude-haiku-4-5-20251001	1/5	Directing and signaling workers requires real-time physical presence, situational awareness, and responsive communication with humans in a dynamic, safety-critical environment. Current AI systems lack the embodied presence and immediate responsive capability to perform this task end-to-end.
Task automatability	claude-sonnet-5	1/5	This involves real-time physical presence, verbal/radio communication, and situational judgment on a plant floor coordinating with other workers tending equipment; current AI cannot perceive plant conditions and direct human workers in this physical, dynamic context.
Adoption barriers	claude-haiku-4-5-20251001	5/5	OSHA and industrial safety regulations typically require qualified human supervisors to direct workers on safety-critical tasks in gas plants. The human-contact and supervisory-sign-off requirements create hard legal barriers to automation.
Adoption barriers	claude-sonnet-5	4/5	Safety-critical gas plant operations typically require certified operators with authority and accountability for directing personnel, and liability for miscommunication in hazardous environments creates strong resistance to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of deploying automated systems (robotics, remote monitoring, communication infrastructure) to replace real-time worker direction would exceed the loaded wage of a human operator performing these signals and directives.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this supervisory/communication function, so any attempted AI solution would require costly sensor networks, robotics, and integration exceeding human labor costs for this narrow task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs real-time worker direction and signaling in industrial settings. This task fundamentally requires on-site human-to-human communication and adaptive coordination that current AI tools cannot execute.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product directs or signals human workers tending auxiliary industrial equipment in gas plants; this remains outside the scope of commercial AI systems.

Operate construction equipment to install and maintain gas distribution systems.

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

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	The construction and gas utility sectors adopt automation slowly; field work remains predominantly labor-intensive, equipment operation is highly specialized, and regulatory/safety requirements create friction against rapid displacement.
Sector adoption velocity	claude-sonnet-5	1/5	Utility and construction sectors are among the slowest to adopt AI/robotics for physical fieldwork, with heavy equipment operation still overwhelmingly manual.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with route planning, equipment diagnostics, or compliance documentation, but equipment operation itself is a hands-on sensorimotor task where meaningful augmentation (e.g., AI co-piloting heavy machinery in real-time) remains largely experimental.
Augmentation potential	claude-sonnet-5	2/5	AI can assist with planning, mapping, and diagnostics (e.g., leak detection, route optimization) but offers little direct assistance for the physical equipment operation itself.
Task automatability	claude-haiku-4-5-20251001	1/5	Operating physical construction equipment in unstructured outdoor/underground environments requires sensorimotor skills, real-time spatial reasoning, and equipment control that current AI systems cannot reliably execute end-to-end; this task involves heavy machinery operation where safety margins and failure costs are prohibitively high.
Task automatability	claude-sonnet-5	1/5	Physical operation of heavy construction equipment for installing and maintaining gas lines requires manual dexterity, real-world spatial judgment, and site-specific adaptation that current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Gas distribution work is tightly regulated by federal, state, and local authorities; operators must hold specific licenses (often Class A Commercial Driver's License or equivalent), and liability for errors (gas leaks, explosions, safety violations) creates legal requirements that a licensed human must perform or directly supervise the work.
Adoption barriers	claude-sonnet-5	5/5	Gas distribution work is heavily regulated, requires licensed/certified operators, safety inspections, and carries high liability for explosions or leaks, mandating human accountability.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized construction equipment, operator licenses, safety oversight, and liability costs mean that autonomous systems capable of this work would be extremely expensive; the hourly cost of deploying such systems would far exceed the loaded wage of a skilled equipment operator.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this physical task, so any hypothetical automation would require expensive robotics and remain costlier than a human operator today.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI system today can autonomously operate construction equipment for gas system installation and maintenance; while some remote robotics and automated machinery exist, they are not general-purpose solutions applicable to the varied conditions of gas distribution work.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously operates construction equipment for gas distribution installation; autonomous construction equipment remains research/pilot stage even for simpler tasks.

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.