Pump Operators, Except Wellhead Pumpers
53-7072.00Median wage $61,770/yr9,780 employed (US)Rank #480 of 923 scored · top 52% by substitution
Tend, control, or operate power-driven, stationary, or portable pumps and manifold systems to transfer gases, oil, other liquids, slurries, or powdered materials to and from various vessels and processes.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution28
Exposure26
Augmentation48
Substitution — the headline: capability discounted by cost, barriers and adoption.
Exposure — technical capability alone, regardless of whether anyone deploys it.
Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.
Tasks on the substitution scale
14 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
7%
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%26
panel mean rating 2.0/5 → substitution pressure 26/100
Technical feasibility todayw 20%26
panel mean rating 2.0/5 → substitution pressure 26/100
Cost vs. human wagew 15%25
panel mean rating 2.0/5 → substitution pressure 25/100
Adoption barriersw 20%inverted — strong barriers lower the score37
panel mean rating 3.5/5 (barrier strength) → substitution pressure 37/100
Sector adoption velocityw 10%24
panel mean rating 1.9/5 → substitution pressure 24/100
Task breakdown (14 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.
Record operating data such as products and quantities pumped, stocks used, gauging results, and operating times.
82CI 72–92 · exposure 87 · augmentation 75 · importance 4.4/5 · click for rater detail
Record operating data such as products and quantities pumped, stocks used, gauging results, and operating times.
82| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Oil, gas, water utilities, and chemical processing—major employers of pump operators—are rapidly adopting automated monitoring and SCADA systems; industrial IoT and digital twins are mainstream in production settings across these sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Industrial and energy sectors are steadily digitizing with SCADA/IoT adoption, but this is slower than software-native sectors due to capital cycles and legacy infrastructure. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Even where humans remain in the loop, AI-powered dashboards and anomaly detection provide real-time alerts and summaries, significantly amplifying an operator's ability to monitor and respond to trends without manual data compilation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enabled dashboards and anomaly detection significantly ease the operator's data recording and cross-checking burden, letting them focus on exceptions and oversight rather than manual transcription. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Recording numerical data (quantities, gauging results, operating times) is purely clerical and highly automatable. Current IoT sensors, PLCs, and SCADA systems already capture this data electronically; AI can parse, validate, and log it with minimal intervention, easily meeting the ≥50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 4/5 | Recording standardized operational data (products, quantities, times) is a structured data-logging task that can largely be handled by sensor integration, SCADA systems, and automated data capture, though initial sensor readings and physical verification still require some human presence. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While some facilities may require human verification or sign-off on critical operating logs for regulatory/compliance reasons, the core recording task itself has few legal or institutional barriers to automation; many modern facilities already bypass manual logging entirely. |
| Adoption barriers | claude-sonnet-5 | 2/5 | There's no licensing requirement for recording data itself, though some regulated industries (e.g., petroleum) may require certified metering or audit trails, creating minor compliance friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Once automated sensor and logging infrastructure is in place (a capital cost amortized across many tasks), the marginal cost of recording additional data points approaches zero, orders of magnitude cheaper than a human operator manually logging each entry. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Once sensors and telemetry are installed, automated data logging is vastly cheaper per data-point than manual recording, though upfront instrumentation and integration costs are non-trivial capital investments. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Industrial facilities widely deploy automated data logging, sensor integration, and SCADA systems in production today. These systems reliably capture and record pump operating metrics without human entry, demonstrating mature, scaled deployments across energy and utilities sectors. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | SCADA and industrial IoT systems already automatically log flow rates, tank levels, and pump run-times in many oil, gas, and chemical facilities today, though older or smaller facilities still rely on manual logs. |
Monitor gauges and flowmeters and inspect equipment to ensure that tank levels, temperatures, chemical amounts, and pressures are at specified levels, reporting abnormalities as necessary.
39CI 25–54 · exposure 42 · augmentation 75 · importance 4.5/5 · click for rater detail
Monitor gauges and flowmeters and inspect equipment to ensure that tank levels, temperatures, chemical amounts, and pressures are at specified levels, reporting abnormalities as necessary.
39| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Large industrial facilities (oil, chemical, water utilities) are widely deploying SCADA and remote monitoring, but adoption is uneven—smaller plants lag significantly, and full autonomous monitoring remains in pilot phase at most sites; displacement of pump operators themselves is limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Oil, gas, and chemical processing industries adopt monitoring automation slowly due to legacy infrastructure, safety certification requirements, and capital cycles, though some large plants use predictive analytics. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-driven monitoring dashboards and alerting systems substantially reduce the cognitive load of manual gauge reading and accelerate anomaly detection, allowing operators to focus on inspection, troubleshooting, and response; this is transformative assistance while keeping the human in the loop for judgment and compliance. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based predictive analytics and anomaly detection significantly help operators catch trends and abnormalities earlier, improving situational awareness while humans remain responsible for physical checks and response. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI systems can reliably monitor digital gauges and flowmeters via SCADA integration and detect anomalies through threshold-based alerts, but physical equipment inspection (corrosion, leaks, wear) and contextual judgment about whether to report require human presence and visual assessment, reducing full automation to roughly half the task. |
| Task automatability | claude-sonnet-5 | 2/5 | Sensor monitoring can be partially automated via SCADA/IoT alarm systems, but physical inspection of equipment and contextual judgment about abnormalities still requires human presence and interpretation on-site.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Industrial facilities face significant regulatory requirements (EPA, OSHA, process safety) and liability concerns if automated systems fail to detect critical anomalies; moreover, many plants require licensed or certified operators to physically inspect equipment and sign off on safety-critical conditions, creating legal and contractual barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Process safety regulations (OSHA, EPA) and liability concerns around chemical/pressure systems typically require a qualified operator to monitor and respond to abnormalities, creating strong regulatory and safety-driven barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | SCADA integration plus AI anomaly detection costs are typically one-tenth to one-fifth of a pump operator's loaded wage, and inference costs are negligible once systems are deployed; however, integration and ongoing oversight add costs that prevent a full order-of-magnitude advantage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Sensor/SCADA systems are cost-effective for continuous data logging, but integrating AI analytics plus maintaining human oversight for physical inspection keeps overall cost comparable to human labor in most facilities. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Industrial monitoring software with AI-driven anomaly detection is deployed in production across refineries, chemical plants, and water treatment facilities; however, physical inspection components and the need for human sign-off on critical abnormalities prevent a fully mature end-to-end automation claim. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Industrial monitoring/alarm software is deployed widely for gauge readings, but AI-driven anomaly detection combined with physical equipment inspection is not a mature end-to-end product in most pump facilities. |
Turn valves and start pumps to start or regulate flows of substances such as gases, liquids, slurries, or powdered materials.
29CI 25–32 · exposure 25 · augmentation 38 · importance 4.3/5 · click for rater detail
Turn valves and start pumps to start or regulate flows of substances such as gases, liquids, slurries, or powdered materials.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Large process plants and refineries have adopted automated and remote SCADA systems for decades, but these are conventional industrial controls, not AI agents. Smaller pump facilities and variable-duty operations remain largely manual; AI adoption in this domain is still in early pilots rather than mainstream production displacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Process industries adopt automation steadily but this sector is physical/industrial with slower AI-driven adoption compared to information-based sectors; existing automation is often traditional control engineering, not modern AI. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Current AI offers limited real-time assistance to pump operators beyond standard monitoring dashboards and alerts that conventional SCADA already provides. AI could assist with predictive maintenance recommendations, but does not meaningfully augment the core task of valve-and-pump operation initiation and flow regulation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled monitoring and predictive analytics can help operators anticipate flow issues and optimize settings, but the physical valve/pump operation itself sees less direct AI augmentation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While initiating pump sequences and valve operations are physically automatable via programmable controllers and SCADA systems, this task typically requires environmental assessment, safety checks, and real-time decision-making based on irregular conditions that current AI systems handle poorly without domain-specific engineering. Partial automation of routine starts exists, but the full task with equal quality across variable conditions does not meet the 50% time-saving bar. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical valve turning and pump starting requires either automated control hardware retrofits (SCADA/PLC) or physical presence; AI software alone cannot perform the physical actuation without existing automation infrastructure already in place.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pump operation in process industries is subject to strict safety regulations, environmental permits, and liability requirements; many jurisdictions require a licensed operator or certified technician to approve or oversee critical valve and pump starts. Legal and certification barriers are substantial, protecting human oversight. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Safety regulations, hazardous materials handling protocols, and liability concerns around industrial accidents create meaningful barriers, though not always requiring a specific human license as with wellhead pumpers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Retrofitting AI-driven automation into existing pump systems involves significant integration, sensor upgrade, and validation costs that often exceed the loaded wage of a single pump operator, especially in smaller or older installations. The capital expenditure and ongoing maintenance offset labor savings for many operations. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Retrofitting facilities with automated valve actuators, sensors, and control systems requires substantial capital investment that may exceed the wage savings for many operations, especially smaller sites. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Deployed industrial automation systems can perform routine valve/pump operations in controlled environments, but these are hard-coded mechanical controllers or conventional PLC systems, not AI-driven. General-purpose AI agents lack the integration depth, real-time sensor fusion, and safety certification required for production deployment in hazardous pump environments. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Many plants already use SCADA/DCS automation for flow regulation, but this is industrial automation/control systems rather than AI per se, and many facilities still rely on manual valve operation, especially older or smaller sites. |
Tend vessels that store substances such as gases, liquids, slurries, or powdered materials, checking levels of substances by using calibrated rods or by reading mercury gauges and tank charts.
28CI 25–30 · exposure 25 · augmentation 50 · importance 4.2/5 · click for rater detail
Tend vessels that store substances such as gases, liquids, slurries, or powdered materials, checking levels of substances by using calibrated rods or by reading mercury gauges and tank charts.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Pump operation in utilities, chemical plants, and petroleum sectors shows slow AI adoption compared to information sectors. Most facilities rely on legacy manual monitoring or basic digital sensors rather than AI-augmented visual inspection systems in production. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Industrial/physical process industries (oil, gas, chemical storage) adopt automation more slowly than office-based sectors, though sensor-based monitoring is a known long-term trend rather than a fast-moving AI adoption pattern. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted computer vision for gauge transcription and anomaly flagging could meaningfully support operator decision-making by accelerating data collection and highlighting unusual readings. However, the core task's physical and judgment components limit the depth of augmentation possible. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled sensor analytics and predictive alerts can help operators monitor trends and catch anomalies more efficiently, augmenting the manual checking process even if it doesn't replace physical verification entirely. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While reading gauges and charts could be partially automated with computer vision, the task requires tending vessels in physical environments with frequent manual intervention and judgment calls about substance levels. Current AI cannot reliably perform the full monitoring and response cycle end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Level monitoring can be automated via sensors and telemetry, but the task as described involves manual physical checks (calibrated rods, gauges) that require on-site presence and physical interaction, limiting full end-to-end automation without capital investment in new instrumentation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Industrial safety regulations (EPA, OSHA) and process control standards often require documented human accountability for vessel monitoring and hazardous substance handling. Liability for spills or equipment failure creates legal pressure for licensed human operators to retain direct oversight and sign-off responsibility. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Safety regulations in industrial/chemical storage settings often require human verification of certain readings, and liability concerns around spills, leaks, or explosions create moderate barriers to full automation of monitoring duties. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Retrofitting AI vision systems for gauge reading, combined with integration and oversight costs, likely exceeds the loaded wage of pump operators in most operational contexts. The infrastructure investment and maintenance overhead add significant expense relative to human monitoring labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Retrofitting facilities with automated sensors, IoT telemetry, and monitoring software involves significant capital and integration costs compared to a human already performing manual checks as part of a broader operator role. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Computer vision systems can identify gauge readings in controlled lab settings, but deployed products for continuous tank monitoring in industrial environments face challenges with varied lighting, gauge styles, and real-time integration with control systems. Existing industrial automation focuses on sensor networks rather than visual monitoring of legacy gauges. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated tank level monitoring systems exist and are deployed in many industrial settings, but this task explicitly describes manual gauge/rod-based checking, which is still common in many facilities and not yet fully replaced by AI-driven systems in this specific job context. |
Read operating schedules or instructions or receive verbal orders to determine amounts to be pumped.
28CI 25–30 · exposure 25 · augmentation 50 · importance 4.0/5 · click for rater detail
Read operating schedules or instructions or receive verbal orders to determine amounts to be pumped.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Pump operators work in manufacturing, utilities, and mining—moderate digitization sectors where automation pilots exist but full deployment remains limited due to safety-critical nature and regulatory oversight requirements. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Oil/gas and industrial processing sectors adopt automation steadily but are not fast digitizers like finance or software; SCADA automation exists but full instruction-interpretation-to-execution AI is still uncommon. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by summarizing operating schedules, transcribing and clarifying verbal orders, and flagging potential anomalies in instructions, improving operator situational awareness without replacing human decision-making on actual pump amounts. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based scheduling assistants and voice-to-text/order parsing tools can help operators quickly interpret instructions and cross check volumes, improving efficiency without replacing human oversight. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task involves reading schedules, instructions, or receiving verbal orders—all interpretable by current AI—but the context-dependent judgment of 'amounts to be pumped' requires real-time operational knowledge and safety validation that would require significant setup and oversight to automate reliably. |
| Task automatability | claude-sonnet-5 | 2/5 | Reading a schedule and executing a pumping instruction is mostly a physical-industrial-control task tied to plant systems; while parsing text/verbal orders is AI-feasible, the task is embedded in real-time plant operations that current off-the-shelf AI cannot fully execute end-to-end at equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pump operation in industrial and municipal settings is often governed by regulatory compliance, safety protocols, and operational liability; any automation must maintain human accountability and may require licensed operators to verify or sign off on pumping decisions, creating legal and organizational barriers. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for this specific task, but safety regulations, plant certification standards, and liability concerns around misjudged pumping volumes create meaningful friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The inference cost of reading/transcribing instructions is low, but integrating with industrial control systems, adding safety oversight, and maintaining human-in-the-loop validation for critical operational decisions makes the all-in cost comparable to or slightly higher than a single pump operator's wage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Industrial control integration, sensors, and safety-certified automation systems require significant capital and maintenance costs, making all-in AI costs not clearly cheaper than an operator's wage for this narrow task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can parse written instructions and transcribe verbal orders, deployed systems lack reliable integration with real-time pump control systems, operational context awareness, and safety verification needed for production deployment in industrial pump operation. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | SCADA/DCS systems already interpret schedules digitally, but deployed AI products that reliably take verbal orders and autonomously determine pumping amounts in production industrial settings are narrow and not widespread. |
Pump two or more materials into one tank to blend mixtures.
28CI 25–30 · exposure 25 · augmentation 50 · importance 3.9/5 · click for rater detail
Pump two or more materials into one tank to blend mixtures.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Most blending operations in industrial settings already use dedicated automation (PLCs, SCADA) installed decades ago; facilities are not rapidly adopting general-purpose AI agents. Migration from legacy control systems to AI is slow and capital-intensive, confined to greenfield projects or major upgrades. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Process manufacturing and chemical/petroleum sectors adopt automation steadily but cautiously due to safety and capital constraints, with AI-specific adoption (versus traditional PLC control) still nascent. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by predicting optimal blend ratios, flagging sensor anomalies, or recommending process adjustments, which would aid human operators in monitoring and decision-making. However, the core manual control task itself (opening valves, adjusting flows) is already largely delegated to hardware automation, limiting the scope of human-AI collaboration. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based predictive analytics and monitoring dashboards can help operators optimize blend ratios and detect anomalies, improving decision-making even though the physical pumping remains human- or automation-controlled. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | The task requires precise control of multiple material flows and real-time blending monitoring, which modern PLC and SCADA systems already handle better than general AI. While AI could theoretically optimize blend ratios, current systems lack reliable sensorimotor control for actual pump valve operation in production environments without significant custom integration. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires physical control of pumps, valves, and tanks with sensor monitoring and hands-on adjustment, which off-the-shelf AI cannot perform end-to-end without extensive industrial automation hardware already in place.4 Only the decision/monitoring layer is AI-automatable, not the physical execution. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Chemical and fluid handling operations typically require licensed operators and carry liability exposure for blend errors (contamination, safety hazards). Regulatory frameworks in petrochemicals and pharmaceuticals often mandate human accountability and certification for critical blending operations. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Safety regulations, hazardous materials handling requirements, and plant liability create moderate barriers; some facilities require certified operators to oversee blending of chemicals or fuels for safety compliance. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The infrastructure cost of retrofitting pump systems with AI-driven control would likely exceed the modest labor cost of an operator monitoring and adjusting a well-maintained blending system. Existing automation (PLC-based) is already cheaper than both human and general-purpose AI solutions. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Retrofitting or building automated blending systems with sensors, actuators, and control software requires significant capital investment, so the all-in cost is not clearly cheaper than an operator, especially at smaller facilities. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Automated blending systems exist in industrial plants, but they rely on purpose-built control hardware (programmable logic controllers, sensors) rather than general AI. Deploying a vision-based or language-based AI agent to physically operate pump controls is not a demonstrated production capability in real facilities. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Industrial control systems and SCADA/DCS with recipe-based blending exist and are mature, but these are traditional automation/PLC systems rather than AI products, and fully autonomous AI-driven blending across diverse plants is not widely deployed. |
Test materials and solutions, using testing equipment.
28CI 25–30 · exposure 25 · augmentation 50 · importance 3.8/5 · click for rater detail
Test materials and solutions, using testing equipment.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Pump operations are largely in industrial, energy, and water management sectors with moderate digitization; while some sites use automated sensors, adoption of AI-driven autonomous testing is slow and limited to data collection rather than decision-making. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Industrial/manufacturing and utility sectors where pump operators work show slower AI adoption relative to information and professional services sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted tools such as automated sensor data visualization, anomaly detection alerts, and predictive maintenance recommendations can meaningfully enhance a pump operator's testing productivity and decision-making, though the human remains the primary actor. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled sensors and data analysis tools can help operators monitor trends and flag anomalies, improving efficiency of interpreting test results even though physical testing remains manual. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Some aspects of testing (running standard equipment, recording measurements) could be partially automated with sensors and software, but interpretation of anomalies, troubleshooting equipment failure, and adapting test parameters to real-world conditions require human judgment. Full end-to-end automation with 50% time savings at equal quality is not achievable with current systems. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical sampling and use of testing equipment on materials/solutions requires manual handling and calibration that current AI cannot perform end-to-end; only data interpretation portions could be automated.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Testing in industrial pump operations is often subject to regulatory compliance (environmental, safety, quality standards), and results must be certified or signed off by qualified personnel. Liability for faulty test results creates strong institutional and legal barriers to full automation without human authorization. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for this specific task, but safety protocols, equipment certification, and plant liability create moderate organizational friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Initial hardware investment in automated testing equipment and integration is substantial, and ongoing maintenance and calibration are costly. For routine pump testing, the all-in cost of AI-driven automation remains comparable to or exceeds the labor cost of a pump operator performing tests. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Automated inline sensors and analyzers have upfront capital and maintenance costs that are often comparable to or higher than human labor for this discrete task in many facilities. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While automated sensors and data logging exist in industrial settings, no deployed end-to-end system reliably performs material/solution testing with the adaptability and judgment required in pump operations without significant human oversight and intervention. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some sensor-based automated testing and monitoring systems exist in industrial plants, but they are narrow-scope and still require human operators to run, calibrate, and interpret results. |
Plan movement of products through lines to processing, storage, and shipping units, using knowledge of interconnections and capacities of pipelines, valve manifolds, pumps, and tankage.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.4/5 · click for rater detail
Plan movement of products through lines to processing, storage, and shipping units, using knowledge of interconnections and capacities of pipelines, valve manifolds, pumps, and tankage.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption is slow and limited to large integrated operations with significant capital investment in advanced SCADA; most pump operators work in traditional refinery and chemical plant settings where legacy systems and regulatory conservatism slow AI deployment. Pilot projects exist but production displacement remains minimal. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Oil and gas/pipeline industries are historically slow adopters of AI-driven automation for core operational decisions, with digitization focused more on monitoring than autonomous planning. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI systems can assist by analyzing historical flow patterns, predicting bottlenecks, and recommending optimal routing, which helps operators plan more efficiently. However, the high stakes and need for human judgment in safety-critical environments limit the transformative potential of augmentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based decision support and optimization tools can meaningfully assist operators in visualizing capacity constraints and suggesting routing options, improving efficiency while humans retain final control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can analyze pipeline data and simulate flow scenarios, end-to-end planning requires real-time decision-making, dynamic rerouting, and integration with on-site manual valve operations and safety overrides that remain operator-dependent. Current systems lack the sensorimotor integration to execute the full task without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | While planning logic could be partially modeled, this requires real-time integration with physical pipeline systems, sensor data, and dynamic operational constraints that current off-the-shelf AI cannot fully manage end-to-end without significant custom engineering. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Environmental regulations, safety certifications, liability requirements, and union agreements in many refineries and chemical plants mandate that licensed operators remain responsible for critical routing decisions. Many jurisdictions require a licensed human to sign off on product movement plans. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Pipeline operations are subject to significant safety regulations, liability concerns for spills/explosions, and often require certified operators, creating substantial barriers to full automation of planning decisions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Implementing and maintaining AI planning systems for specialized pipeline operations requires significant domain expertise and custom integration, while pump operators have moderate wages; the all-in cost of AI systems currently exceeds the cost of human operators for this niche task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Custom industrial control and optimization systems require significant integration, sensor infrastructure, and ongoing oversight costs that make them comparable to or more expensive than experienced human operators in many facilities. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | SCADA systems and industrial control software exist for monitoring and logging, but no deployed product reliably plans and executes complex multi-unit product movement autonomously. Existing tools require human operators to validate plans and manually intervene; they are primarily supervisory, not autonomous planners. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some pipeline scheduling software and optimization tools exist, but fully autonomous planning of product movement through complex interconnected pipeline networks is not reliably deployed as a standalone AI product replacing human operators. |
Add chemicals and solutions to tanks to ensure that specifications are met.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.0/5 · click for rater detail
Add chemicals and solutions to tanks to ensure that specifications are met.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of chemical automation in pump operations remains slow outside large-scale refineries and utilities; most small and mid-size industrial facilities still rely on manual operator oversight and decision-making. Sectors where pump operators work (utilities, small manufacturing) lag in automation technology adoption. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Industrial process operations, particularly in oil/gas and manufacturing pumping operations, show slower AI adoption compared to information-sector work due to physical infrastructure and safety-critical requirements. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Current monitoring systems and sensor alerts can assist operators by flagging out-of-specification conditions and recommending chemical adjustments, improving situational awareness. However, AI assistance is limited by the need for operators to make safety-critical judgment calls and verify tank conditions manually. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based monitoring and predictive systems can assist operators by flagging when chemical levels deviate from specifications or recommending dosing adjustments, improving decision-making even though physical execution remains manual. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While chemical dispensing itself can be partially automated with metering pumps and sensors, the task requires real-time judgment about specification compliance, tank conditions, and troubleshooting deviations—capabilities current AI systems lack without significant human oversight. Automating the full end-to-end task with ≥50% time savings at equal quality is not demonstrable with off-the-shelf systems. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a physical task requiring manual handling, dosing, and verification of chemicals in industrial tanks, which current AI cannot perform end-to-end without robotic hardware and sensor integration.atibility.The cognitive component (determining amounts based on specs) could be automated, but the physical addition cannot be by AI alone.rated |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory compliance (EPA, OSHA, process safety management) and liability for incorrect chemical addition create strong organizational and legal barriers to full automation. Most industrial settings require licensed or trained human operators to verify and sign off on chemical additions to meet safety and quality standards. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Chemical handling in industrial settings is often subject to safety regulations, environmental compliance, and certification requirements, and errors in chemical concentration can have serious safety/environmental consequences requiring human accountability. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automated chemical injection systems are capital-intensive and require specialized maintenance, calibration, and oversight infrastructure. For low-volume or variable chemical operations, the installed and operating costs often remain comparable to or exceed the cost of a pump operator's labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Deploying automated chemical dosing systems requires significant capital investment in sensors, actuators, and control integration, making all-in costs comparable to or higher than human labor for many facilities. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Chemical dosing systems and automated dispensers exist in industrial settings, but they typically operate within narrow, pre-programmed parameters and require human operators to verify compliance, adjust for variable tank conditions, and handle exceptions. Production systems are heavily human-supervised and do not perform the full task reliably without intervention. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated dosing systems exist in some process industries but are engineering/control-system solutions rather than AI products, and general-purpose AI models cannot physically perform this task in production today. |
Communicate with other workers, using signals, radios, or telephones, to start and stop flows of materials or substances.
22CI 19–25 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail
Communicate with other workers, using signals, radios, or telephones, to start and stop flows of materials or substances.
22| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Pump operators work in traditional industrial sectors (utilities, refineries, manufacturing) with slow digital transformation and conservative safety cultures. Adoption of AI for critical operational coordination in these sectors remains minimal and mostly confined to monitoring rather than autonomous control. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Industrial/manufacturing plant operations are historically slower to adopt AI-driven coordination tools compared to information-sector work, with automation focused on control systems rather than communication tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by logging communications, generating alerts, or scheduling coordination between workers, modestly raising operator productivity. However, the core task—real-time negotiated communication to manage material flows—remains human-centric, limiting augmentation impact. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled monitoring, alerting systems, and radio/communication logging can support operators by providing real-time data and flagging anomalies, improving decision-making during flow control tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could theoretically generate or interpret signals and manage radio/telephone coordination, the task requires real-time, safety-critical decision-making with other workers in a physical industrial environment. Current systems cannot reliably replace the human judgment and situational awareness needed to coordinate equipment start/stop safely. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires physical presence, real-time situational awareness of equipment/material flow, and coordination with other on-site workers, which current AI cannot perform end-to-end though some monitoring/alerting could be automated. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Safety regulations, OSHA oversight, and operational liability create strong barriers: someone must be legally accountable for material flow in hazardous environments. Industrial facilities typically require human sign-off on pump operations and inter-worker coordination for compliance and risk management. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Safety-critical industrial operations often have regulatory oversight, liability concerns for material spills/hazards, and require certified operators to be present or in the loop for such communications. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI systems to manage industrial communication networks, including integration and liability, exceeds the wage of a pump operator. A single operator's labor is lower-cost than the infrastructure and oversight required for autonomous coordination. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While sensor-based automation exists, replacing the communication and judgment role with AI still requires significant hardware/integration investment comparable to or exceeding wages for this narrow task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform this task end-to-end; communication in industrial settings remains dependent on human operators. AI can assist with scheduling or logging communications, but cannot yet independently coordinate multi-worker material flow operations on production systems. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | SCADA and automated control systems exist for flow monitoring, but human-to-human communication for coordination during physical pump operations remains manual in most industrial settings today. |
Tend auxiliary equipment such as water treatment and refrigeration units, and heat exchangers.
19CI 7–30 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail
Tend auxiliary equipment such as water treatment and refrigeration units, and heat exchangers.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Industrial plants adopt remote monitoring and some automated controls, but these sectors (water utilities, petrochemical, HVAC) are slow to adopt fully autonomous tending systems due to safety requirements, capital constraints in legacy infrastructure, and preference for human oversight of critical equipment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Industrial and utility operations sectors have low digitization of physical tending tasks, with AI adoption concentrated in monitoring/analytics rather than physical task execution. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-driven monitoring systems and predictive maintenance alerts meaningfully assist operators by flagging anomalies and scheduling maintenance, raising their ability to respond proactively. However, the core physical work of adjustment and repair remains human-centered. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-powered sensor analytics and predictive maintenance systems can help operators monitor equipment health and flag issues, improving decision-making even though physical tending remains manual. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While monitoring equipment performance through sensors is automatable, tending auxiliary systems requires hands-on intervention (adjusting valves, replacing filters, responding to equipment failures) that current AI cannot perform without physical embodiment. Only data-collection and alert components are readily automatable. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical monitoring and maintenance task involving auxiliary equipment that requires physical presence, manual adjustments, and sensory inspection that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Safety-critical equipment in industrial settings (water treatment, refrigeration, heat exchangers) typically requires licensed operators or certified technicians to perform maintenance and adjust systems, and liability for equipment failure or safety incidents creates strong legal and regulatory barriers to unmanned automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not strictly licensed work, industrial safety protocols, equipment liability, and the need for physical intervention in emergencies create meaningful organizational and safety barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current industrial automation for this task (sensors, controllers, monitoring software) costs substantially because physical intervention and human expertise still drive most tending work. Full substitution would require robotics, which remains expensive relative to the typical pump operator wage. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Physical tending requires robotic actuation and on-site presence that AI systems cannot provide cheaply; equipping facilities with robotics for this narrow task would cost far more than a human operator. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | SCADA and industrial control systems can monitor equipment automatically, but no deployed AI system reliably performs the full range of physical tending tasks (maintenance, repairs, adjustments) required. Existing automation is narrow in scope and requires human intervention. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product physically tends water treatment units, refrigeration units, or heat exchangers; this remains firmly in the domain of human physical operators supported at most by sensor monitoring software. |
Collect and deliver sample solutions for laboratory analysis.
16CI 14–19 · exposure 16 · augmentation 25 · importance 4.0/5 · click for rater detail
Collect and deliver sample solutions for laboratory analysis.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Pump operations remain in capital-intensive, physical, and safety-regulated sectors with slow digital transformation. Current adoption of autonomous sample collection in these environments is negligible; organizations continue relying on human operators due to safety, liability, and technical maturity concerns. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Industrial/manufacturing plant operations are a low-digitization, physical-labor-heavy sector with minimal AI/robotic adoption for tasks like manual sample collection. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by recommending sampling schedules or flagging when samples should be collected based on process parameters, but the physical act of collection and the judgment around contamination prevention benefit minimally from current AI assistance. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with scheduling sample collection, logging results, or flagging anomalies in lab data, but offers little help with the core physical collection and delivery task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Sample collection and delivery involves physical manipulation in potentially hazardous industrial environments, requiring contextual judgment about which samples to take and when. While logistics routing could be partially automated, the core act of collecting solutions safely and correctly remains heavily dependent on human presence and tactile decision-making. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical sample collection and transport from field/plant equipment requires manual manipulation of valves, containers, and mobility in industrial settings, which current AI/robotics cannot reliably perform end-to-end.subst |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Significant regulatory and liability barriers apply: environmental sampling must often meet regulatory chain-of-custody requirements, sample integrity standards, and safety protocols that legally require human validation or signature. Industrial facilities impose strict access controls and safety training mandates. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not licensed work per se, safety protocols, chain-of-custody requirements for lab samples, and physical plant access create meaningful organizational and procedural friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Deploying autonomous systems capable of safe sample handling in industrial settings (robotic arms, mobile platforms, sensing) would be prohibitively expensive compared to the loaded wage of a pump operator performing occasional sampling. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI system performing physical sampling and delivery, so the human remains the only cost-effective option for this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No current off-the-shelf AI system reliably performs end-to-end sample collection and physical delivery in industrial pump operations. This requires mobile manipulation in unstructured environments with safety and contamination risks that exceed deployed autonomous capabilities today. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously collects physical liquid/gas samples from pumps and delivers them to a lab; this remains a manual task performed by human operators. |
Connect hoses and pipelines to pumps and vessels prior to material transfer, using hand tools.
7CI 5–10 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail
Connect hoses and pipelines to pumps and vessels prior to material transfer, using hand tools.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of AI/robotics in pump operation is minimal; the sector remains traditional and decentralized across small to mid-sized facilities with limited digital infrastructure and high costs for automation retrofitting. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Pump operation in industrial/chemical/oil sectors involves heavy physical infrastructure with low digitization of manual connection tasks and minimal AI/robotics adoption for this specific manual step. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI offers minimal assistance; perhaps planning tools or checklists could be augmented with AI recommendations, but the core physical task of connecting hoses leaves little room for meaningful AI enhancement without a human doing the hands-on work. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could potentially assist with checklists, safety verification, or diagnostic monitoring around the process, but offers little direct assistance to the physical act of connecting hoses and pipelines. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task involves physical manipulation in the field requiring dexterity, spatial reasoning, and real-time problem-solving with hand tools. Current AI systems cannot perform hands-on mechanical assembly and connection work in real industrial environments. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manual task requiring hand-eye coordination, dexterity, and hand tools to connect hoses and pipelines; no current AI system can perform this physical manipulation end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Significant barriers exist: liability concerns around incorrect connections in hazardous environments (pressurized systems, flammable materials), regulatory oversight of industrial piping safety, and the requirement for a qualified human operator to verify and sign off on safety-critical connections. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While no formal licensing typically governs this specific step, safety protocols, liability for spills/leaks during hazardous material transfer, and physical presence requirements create meaningful friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Specialized robotics and AI systems capable of this task would require substantial capital investment, custom integration, and maintenance, far exceeding the loaded cost of a skilled pump operator performing the work. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Robotic systems capable of this dexterous physical task would require expensive specialized hardware and integration far exceeding the cost of a human operator performing this quick task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI products can autonomously connect hoses and pipelines to pumps and vessels using hand tools in production settings. Robotic systems capable of this work exist but are custom, not off-the-shelf AI solutions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed products perform physical hose/pipeline connection using hand tools; robotic manipulation for such variable industrial connections remains research-stage at best. |
Clean, lubricate, and repair pumps and vessels, using hand tools and equipment.
7CI 5–10 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail
Clean, lubricate, and repair pumps and vessels, using hand tools and equipment.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Pump operation occurs predominantly in physical, industrial, and energy sectors with lower digitization and slower AI adoption rates; no measurable displacement by automation is occurring. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Industrial pump maintenance is a physically demanding, low-digitization field with minimal robotic or AI adoption in production settings. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with maintenance scheduling, diagnostic support, or documentation, but hands-on cleaning, lubrication, and repair remain fundamentally manual tasks where augmentation opportunities are limited and tools remain at the informational level. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with predictive maintenance scheduling or diagnostics via sensor data, but offers little direct help with the hands-on cleaning, lubrication, and repair actions themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation of mechanical equipment in varied, often constrained spaces using hand tools—work that current robotics cannot reliably perform at scale. The dexterity, spatial reasoning, and real-time problem-solving needed for maintenance operations remain far beyond what deployed automation can handle. |
| Task automatability | claude-sonnet-5 | 1/5 | This is physical manual labor involving hand tools on industrial equipment, requiring dexterity, mobility, and situational judgment that current AI systems cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pump maintenance often occurs in regulated industrial/energy facilities with strict safety certifications and legal liability for improper repair. Human sign-off and on-site presence are typically required by industry standards and insurance. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement specifically bars automation, but safety protocols, physical access constraints, and equipment liability create moderate friction against any automated substitute. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Current industrial robotics capable of such maintenance work costs hundreds of thousands to millions of dollars per unit, with significant integration overhead, vastly exceeding the loaded wage of a skilled pump operator. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-driven robotic substitute for this physical maintenance work, so the human remains the only cost-effective option today. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No commercial AI or robotic system currently performs pump cleaning, lubrication, and repair work reliably in production environments. This remains a domain requiring human physical presence and judgment. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously cleans, lubricates, and repairs industrial pumps using hand tools; robotic manipulation for this remains research-stage at best. |
Related occupations — Transportation & Material Moving
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.