Janitors and Cleaners, Except Maids and Housekeeping Cleaners

37-2011.00

Median wage $36,840/yr2,209,760 employed (US)Rank #651 of 923 scored · top 71% by substitution

Keep buildings in clean and orderly condition. Perform heavy cleaning duties, such as cleaning floors, shampooing rugs, washing walls and glass, and removing rubbish. Duties may include tending furnace and boiler, performing routine maintenance activities, notifying management of need for repairs, and cleaning snow or debris from sidewalk.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution21

Exposure12

Augmentation23

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

21 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%13

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

Technical feasibility todayw 20%8

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

Cost vs. human wagew 15%8

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

Adoption barriersw 20%inverted — strong barriers lower the score64

panel mean rating 2.5/5 (barrier strength) → substitution pressure 64/100

Sector adoption velocityw 10%5

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

Task breakdown (21 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.

Requisition supplies or equipment needed for cleaning and maintenance duties.

CI 30–57 · exposure 38 · augmentation 50 · importance 3.7/5 · 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	Janitor and cleaning services are labor-intensive, often operate in smaller organizations with limited digitization, and lag in tech adoption compared to professional services or information sectors. Procurement automation in this sector is developing slowly.
Sector adoption velocity	claude-sonnet-5	2/5	Janitorial services and building maintenance are a low-digitization sector with modest software adoption; automated inventory/requisition tools are used mainly in larger facilities management operations, not widespread.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist by tracking consumption patterns, suggesting reorder times, or auto-populating requisition forms with historical items, meaningfully reducing time spent on routine ordering while a janitor or supervisor retains final decision-making authority.
Augmentation potential	claude-sonnet-5	3/5	AI-based inventory tracking and reorder alerts can meaningfully help janitorial staff or supervisors know when and what to requisition, improving efficiency without replacing the judgment of physical stock checks.
Task automatability	claude-haiku-4-5-20251001	2/5	Requisitioning supplies requires knowing what is needed, inventory levels, and organizational procurement systems. While AI could partially automate supply tracking or suggest reorders based on consumption patterns, the task involves human judgment about which specific products, quantities, and vendors to use, and integration with legacy procurement systems is typically manual or semi-automated at best.
Task automatability	claude-sonnet-5	3/5	Generating requisition requests, checking inventory levels, and placing orders is a text/data task that AI can largely automate, but it requires integration with inventory systems and physical verification of stock.5, but sits well below full automation because physical stock checks remain human.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Most organizations require human authorization or approval of supply orders for budget and accountability reasons. Procurement policies, vendor contracts, and accounting controls create moderate friction that prevents full autonomous substitution.
Adoption barriers	claude-sonnet-5	1/5	No licensing or regulatory requirement restricts who can requisition supplies; it's a routine administrative task with no liability concerns.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The task involves relatively low-cost per-instance labor (a janitor spending minutes to order supplies). Building and deploying AI systems to automate procurement, even with off-the-shelf tools, typically carries setup and integration costs that exceed the direct wage savings for routine requisitions.
Cost vs. human wage	claude-sonnet-5	3/5	Automated procurement/inventory systems are cheap to run once set up, but implementation, integration with existing supply chains, and human oversight for exceptions keep costs roughly comparable to manual ordering for small operations.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some inventory management and procurement software exists with automation features, but most janitor/cleaning operations rely on manual or semi-manual requisition processes. Fully autonomous end-to-end requisitioning without human oversight remains rare in deployed systems, especially for smaller facilities and organizations.
Technical feasibility today	claude-sonnet-5	3/5	Procurement and inventory management software with AI-driven reordering exists and is used in facilities management, but janitorial-specific requisition workflows are often manual or lightly digitized in smaller operations.

Clean building floors by sweeping, mopping, scrubbing, or vacuuming.

CI 35–40 · exposure 25 · augmentation 25 · 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	2/5	Adoption is slow and concentrated in large, high-profile facilities with controlled environments; the majority of janitor work occurs in small-to-medium buildings, schools, and offices where capital investment in robots remains limited.
Sector adoption velocity	claude-sonnet-5	2/5	Cleaning/janitorial services is a low-digitization, physical-labor sector with only scattered adoption of cleaning robots in large retail/warehouse spaces, not broad or fast-moving adoption overall.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI tools offer minimal augmentation to janitors performing floor cleaning—robotic assistance exists but is largely autonomous rather than collaborative, and janitors gain little productivity boost from current AI systems for this manual task.
Augmentation potential	claude-sonnet-5	2/5	Robotic vacuums/scrubbers can offload some routine large-area cleaning, letting workers focus on detail work, but this augmentation is limited to certain floor types and facilities rather than being a general productivity multiplier.
Task automatability	claude-haiku-4-5-20251001	2/5	While robotic vacuum cleaners exist and can handle some floor cleaning autonomously, they cannot yet perform the full range of sweeping, mopping, and scrubbing tasks reliably or achieve 50% time savings compared to human janitors when accounting for setup, edge cases, and diverse floor types.
Task automatability	claude-sonnet-5	2/5	Robotic vacuums and scrubbers exist for large flat commercial floors, but sweeping/mopping edges, stairs, obstacles, and varied surfaces still requires human labor or extensive custom setup, so overall time saving is well below 50% in most buildings.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No legal licensing requirement exists for janitorial work, but organizational friction is moderate: facilities managers prefer established human workers familiar with buildings, and customer perception of cleanliness still favors human oversight and responsiveness.
Adoption barriers	claude-sonnet-5	1/5	No licensing, liability, or regulatory requirement mandates a human perform floor cleaning; it's a low-barrier physical task.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Robotic floor cleaning systems cost tens of thousands of dollars upfront plus maintenance, while a janitor's loaded wage remains lower when amortized across diverse cleaning tasks and buildings; automation is not yet cost-competitive for most janitor roles.
Cost vs. human wage	claude-sonnet-5	2/5	Commercial cleaning robots have high upfront and maintenance costs and still need human supervision and supplementary manual cleaning, making all-in cost per task-equivalent comparable to or higher than low-wage janitorial labor in most settings.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Commercial robotic floor cleaners are deployed in some controlled environments (airports, malls) but have narrow scope, require frequent human intervention, and struggle with obstacles, stairs, and varied floor conditions—falling short of reliable production-scale performance across typical building environments.
Technical feasibility today	claude-sonnet-5	2/5	Autonomous floor-scrubbing robots (e.g., Brain Corp-powered machines) are deployed in some big-box retail and warehouses, but they handle only a narrow subset of floor cleaning and require human oversight, refilling, and edge/detail cleaning.

Notify managers concerning the need for major repairs or additions to building operating systems.

CI 28–39 · exposure 25 · augmentation 38 · 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	2/5	Smart building technologies are growing in large commercial properties, but the specific task of notifying managers about major repairs remains mostly manual. Adoption is concentrated in new or well-resourced facilities; most buildings still rely on janitor reporting, indicating slow and uneven deployment.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial/facilities maintenance is a low-digitization, physical-labor sector with minimal AI agent deployment for this kind of task.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-powered monitoring systems and predictive maintenance alerts can help janitors identify repair needs earlier by flagging equipment anomalies, extending their effectiveness and reducing missed problems. However, the janitor's visual inspection and contextual knowledge remain essential for accurate assessment.
Augmentation potential	claude-sonnet-5	2/5	Simple apps or chatbots could help draft or route a repair notification once identified, but AI does not aid in the visual/physical detection itself.
Task automatability	claude-haiku-4-5-20251001	2/5	AI monitoring systems could detect some sensor data indicating repair needs, but the task requires judgment about severity, cost-benefit analysis, and prioritization that current AI handles only in narrow, pre-defined scenarios. Reliable end-to-end automation with 50%+ time savings requires human interpretation of building conditions and contextual factors.
Task automatability	claude-sonnet-5	2/5	The core work is physical inspection and judgment about building systems, which AI cannot perform; only the notification/reporting portion could be automated, leaving most of the task's value untouched.
Adoption barriers	claude-haiku-4-5-20251001	3/5	While there are no strict legal or licensing barriers to AI monitoring systems, organizational friction exists: building managers often prefer direct human reporting, trust-building with facilities staff, and the fact that janitors already navigate the space, creating natural observation opportunities that automated systems may miss.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement for this specific act, but organizational reliance on human presence and judgment for identifying physical defects creates practical friction.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	Smart building sensors and monitoring software are moderately priced, but integration, tuning, and human oversight for notification decisions makes the all-in cost comparable to paying a janitor to occasionally report obvious problems. Neither has a clear cost advantage.
Cost vs. human wage	claude-sonnet-5	2/5	A human janitor already performs this as a low-cost byproduct of routine cleaning; replicating detection with sensors/AI would require capital investment exceeding the marginal cost of a verbal report.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While building automation systems and IoT monitoring exist, none reliably perform the full task of identifying, assessing, and notifying managers about major repairs independently. Current products generate alerts from sensors but require human verification and judgment about what constitutes a 'major' repair needing manager notification.
Technical feasibility today	claude-sonnet-5	2/5	No deployed product autonomously detects need for major repairs and notifies managers on behalf of a janitor; sensor-based building monitoring exists but isn't integrated into this worker's task flow.

Strip, seal, finish, and polish floors.

CI 24–35 · exposure 20 · augmentation 38 · importance 3.7/5 · 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	Adoption of automated floor care equipment in commercial cleaning is slow, concentrated in large facilities with standardized layouts. Most small to mid-size organizations still rely on manual labor due to cost and variability in floor types and layouts.
Sector adoption velocity	claude-sonnet-5	1/5	Building services and janitorial sectors show low digitization and slow adoption of robotics for floor finishing tasks specifically, compared to simple vacuuming.
Augmentation potential	claude-haiku-4-5-20251001	3/5	Power equipment (floor buffers, strippers) and chemical guides can assist janitors in working faster and with less physical strain. AI-guided application or quality monitoring could provide moderate productivity gains, but augmentation remains limited compared to tasks involving decision-making.
Augmentation potential	claude-sonnet-5	2/5	Powered equipment and some semi-autonomous scrubbers assist workers, but there is little AI-specific augmentation of the judgment-heavy stripping/sealing/polishing process.
Task automatability	claude-haiku-4-5-20251001	2/5	Floor finishing requires navigating complex, variable environments with obstacles, precise chemical application, and quality control. While some equipment (e.g., automatic floor strippers) exists, end-to-end automation achieving 50% time savings at equal quality across diverse spaces remains infeasible with current technology.
Task automatability	claude-sonnet-5	2/5	Some autonomous floor-scrubbing machines exist, but stripping, sealing, and finishing floors requires chemical application, edge work, and quality judgment that current robots cannot fully replicate end-to-end.in most facilities.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	Physical presence on customer premises and safety liability for slippery floors create moderate friction. However, no legal licensing requirement exists for floor finishing, and organizations can adopt automation if cost-effective, limiting hard barriers.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but physical environment variability (furniture, edges, stairs, different floor types) and liability for damaged surfaces create moderate practical friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized equipment and materials for floor finishing remain expensive relative to human labor costs. Current robotic or semi-automated solutions require high upfront capital and ongoing maintenance, making them costlier than trained janitors for most facilities.
Cost vs. human wage	claude-sonnet-5	2/5	Specialized floor-care robots have high upfront capital and maintenance costs relative to low-wage janitorial labor, making them cost-comparable or more expensive except at very large scale.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial system reliably performs the full strip-seal-finish-polish cycle autonomously. Robotic floor cleaners handle basic wet cleaning but cannot execute the multi-stage chemical application and finishing work that this task demands.
Technical feasibility today	claude-sonnet-5	2/5	Autonomous scrubber-dryers are deployed in some large commercial settings, but stripping/waxing/finishing floors is rarely automated in production and still relies on human labor for most facilities.

Gather and empty trash.

CI 15–34 · exposure 13 · augmentation 0 · 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	This task occurs predominantly in laggard sectors—small/medium facilities, diverse building types, and organizations with limited capital investment in automation. Adoption of robotic trash collection remains negligible in the broader janitor workforce.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial services are a low-digitization, physically-intensive sector with minimal AI or robotics adoption in production trash collection tasks.
Augmentation potential	claude-haiku-4-5-20251001	1/5	AI does not meaningfully assist a human gathering and emptying trash; the task is manual and repetitive without elements where algorithmic decision-support adds value while a human remains in the loop.
Augmentation potential	claude-sonnet-5	1/5	Current AI offers essentially no assistance to a human performing manual trash gathering and disposal, as the task is purely physical with no cognitive or planning component AI could meaningfully enhance.
Task automatability	claude-haiku-4-5-20251001	2/5	Robotic trash collection systems exist but require significant infrastructure setup, human intervention for contaminated/heavy items, and varied bin locations/designs. Current robots cannot reliably replicate the full task end-to-end at 50% time savings across typical building layouts.
Task automatability	claude-sonnet-5	1/5	Emptying trash requires physical navigation, manipulation of bags/bins, and mobility through varied environments that current AI systems cannot perform; this is a robotics/embodiment problem, not a software automation problem.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few formal licensing barriers exist for trash collection; however, building layouts, safety liability, customer preferences for human workers, and organizational inertia create moderate friction to adoption of automation.
Adoption barriers	claude-sonnet-5	1/5	No licensing, regulatory, or liability barriers specifically restrict automating trash removal; the barrier is purely technical/physical capability, not legal or organizational.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Robotic trash systems (hardware, maintenance, integration, charging infrastructure) are far more expensive than employing janitors at typical wages, especially when accounting for the low-skill, low-wage nature of this work.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-driven solution for this physical task, so any robotic alternative would be far more expensive than human labor including hardware, maintenance, and limited scope of operation.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Autonomous trash collection robots are in limited deployment (e.g., specialized facilities, controlled environments) but are not widely integrated into production janitor workflows. Most deployments remain pilot-stage with substantial human oversight and fallback needs.
Technical feasibility today	claude-sonnet-5	1/5	No deployed products reliably gather and empty trash across varied commercial or institutional settings today; robotic trash handling remains research/pilot stage at best.

Mix water and detergents or acids in containers to prepare cleaning solutions, according to specifications.

CI 24–24 · exposure 16 · 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	Janitorial and cleaning sectors are labor-intensive, low-digitization, fragmented industries with limited capital investment in automation compared to information and financial sectors.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial services are a low-digitization, physically-embodied sector with minimal AI/robotics adoption for granular chemical-handling subtasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist in recommending or documenting correct solution recipes and proportions, but the core physical task of mixing solutions offers minimal augmentation opportunity since the human is already directly manipulating the containers.
Augmentation potential	claude-sonnet-5	2/5	Pre-set dosing dispensers and mixing instructions (sometimes app-guided) offer marginal assistance, but AI does not meaningfully enhance the human's execution of this specific mixing task.
Task automatability	claude-haiku-4-5-20251001	2/5	While mixing solutions involves repeatable steps, current AI lacks the embodied capability to physically handle containers, measure liquids, and perform chemical mixing in real-world environments without human setup and intervention.
Task automatability	claude-sonnet-5	2/5	This requires physical manipulation of containers and precise measurement of chemicals in real-world space, which current general AI systems cannot execute end-to-end without robotic embodiment.atibility. Simple robotic dosing systems exist for industrial settings but are not general-purpose janitorial solutions.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Chemical handling and occupational safety regulations create some friction, though mixing cleaning solutions is not a licensed task; human workers perform it routinely without credentials, leaving moderate barriers to automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement exists for mixing cleaning solutions, though safety concerns around chemical handling (e.g., avoiding acid/bleach mixing errors) create some liability-driven caution around automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of deploying a robot system (purchasing, maintenance, safety infrastructure) far exceeds the wage cost of a human janitor performing this routine task for many years.
Cost vs. human wage	claude-sonnet-5	1/5	Without a mature robotic solution, any automation would require expensive specialized hardware that costs far more than a janitor's time for this small subtask.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs autonomous solution preparation at scale in production facilities; this requires physical manipulation and chemical safety compliance that robotics has not yet matured to handle in general use.
Technical feasibility today	claude-sonnet-5	1/5	No deployed consumer or commercial product autonomously mixes cleaning solutions for janitorial staff today; this remains a manual physical task performed by humans.

Remove snow from sidewalks, driveways, or parking areas, using snowplows, snow blowers, or snow shovels, or spread snow-melting chemicals.

CI 10–35 · exposure 13 · augmentation 25 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Snow removal remains largely manual and labor-intensive across most sectors; adoption of autonomous systems is confined to experimental pilots and a few high-investment municipalities, reflecting slow uptake in a traditionally low-tech sector.
Sector adoption velocity	claude-sonnet-5	1/5	Building maintenance and janitorial services are a low-digitization, physically-oriented sector with minimal AI/robotics adoption for this specific outdoor task.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI-powered equipment (e.g., route optimization, real-time weather integration) offers modest assistance to human operators in planning and execution, but does not fundamentally transform individual worker productivity on the core physical removal task.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with route/schedule optimization or weather-based chemical application timing, but offers minimal direct assistance to the physical act of clearing snow.
Task automatability	claude-haiku-4-5-20251001	2/5	While snow removal equipment exists and autonomous snow plows are in early development, current deployed systems cannot reliably handle the full end-to-end task across variable terrain, weather conditions, and obstacles at 50% time savings. Manual oversight, equipment repositioning, and chemical application still require substantial human judgment and intervention.
Task automatability	claude-sonnet-5	1/5	This is a physical manual/equipment-operated task requiring mobility, perception of terrain, and physical force; no AI system can perform it end-to-end today, only robotic hardware which is not general-purpose or deployed.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Liability concerns for autonomous equipment operating on public sidewalks and driveways, combined with the need for equipment ownership/maintenance decisions and customer preference for human crews, create moderate friction against full automation adoption.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but liability for injury/property damage from equipment, unpredictable outdoor conditions, and need for judgment on chemical application add some friction to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Current autonomous snow removal systems remain capital-intensive and require ongoing technical support, making total cost per cleared area comparable to or higher than hiring seasonal labor, especially for small to medium properties.
Cost vs. human wage	claude-sonnet-5	1/5	Autonomous snow-clearing robotics remain expensive, low-throughput, and unreliable in variable terrain, making a human with a plow or shovel far cheaper currently.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some autonomous or semi-autonomous snow removal equipment exists in limited deployment (e.g., robotic sidewalk clearers), but these operate in narrow, controlled contexts and require human monitoring. No mature production systems reliably replace the full task across diverse real-world environments without significant human supervision.
Technical feasibility today	claude-sonnet-5	1/5	No commercial products autonomously clear snow from sidewalks and lots at scale; autonomous snowplow trials exist only in narrow research/pilot contexts like large parking lots with GPS guidance.

Dust furniture, walls, machines, or equipment.

CI 15–29 · exposure 13 · augmentation 13 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Building services and janitorial work remain fragmented, labor-intensive sectors with limited tech adoption. Most organizations rely on low-cost human labor; cleaning robots have seen negligible production deployment in mainstream commercial or institutional facilities.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial services are a low-digitization, physically dependent sector with minimal AI/robotics adoption for tasks like dusting.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Powered dusting tools and light robotic assists could modestly improve a cleaner's efficiency or reduce physical strain, but current AI systems offer minimal augmentation to the core task of dusting—most productivity gains would come from better tools or ergonomics rather than AI intelligence.
Augmentation potential	claude-sonnet-5	1/5	Current AI tools offer no meaningful assistance to a human performing manual dusting tasks.
Task automatability	claude-haiku-4-5-20251001	2/5	Dusting furniture and equipment requires physical manipulation in varied, unstructured environments with fragile or sensitive items. Current robots can dust in controlled settings, but autonomous systems struggle with obstacle navigation, variable surfaces, and the dexterity needed to avoid damage—falling well short of the 50% time-saving threshold against human cleaners.
Task automatability	claude-sonnet-5	1/5	Dusting requires physical manipulation in varied, unstructured environments; no widely deployed AI/robotic system performs general dusting of furniture, walls, and equipment end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Dusting is a low-skill task with few regulatory barriers and minimal liability concerns for damage—a broom or cloth causes limited harm. However, organizations often value human presence for security and general building maintenance, and customers may prefer visible human care, creating modest friction against full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal barriers prevent automation, but physical/environmental complexity (varied surfaces, obstacles, fragile items) creates practical friction against substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Cleaning robots capable of dusting remain expensive to purchase, deploy, and maintain, while human janitors command modest hourly wages. The upfront capital and integration cost heavily outweigh the savings from a task that occupies only part of a cleaner's day.
Cost vs. human wage	claude-sonnet-5	1/5	Without viable automation, a human cleaner remains cheaper and more effective than any hypothetical robotic solution, which would require expensive custom hardware.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Specialized cleaning robots exist in research and limited commercial deployments, but no mainstream product reliably dusts general furniture and equipment at the quality and speed of human workers. Existing systems have narrow scope and require structured environments, far from production-scale reliability.
Technical feasibility today	claude-sonnet-5	1/5	There are no mature commercial products that reliably dust furniture, walls, or equipment; this remains outside the scope of even advanced cleaning robots which focus mainly on floors.

Service, clean, or supply restrooms.

CI 15–24 · exposure 8 · augmentation 25 · 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	Janitorial work remains concentrated in small, low-digitization facilities, schools, and contract companies with limited capital for robotics investment. Industry adoption of automation is laggard; most restroom servicing is still manual and decentralized across thousands of small operators.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial services are a low-digitization, physical labor sector with minimal AI/robotics adoption in production settings.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI can assist with supply inventory management, scheduling, and quality inspection via computer vision, but offers minimal productivity boost to the core act of physically cleaning and servicing restrooms. The human remains the primary performer of nearly all value-added work.
Augmentation potential	claude-sonnet-5	2/5	Some tools like smart dispensers or scheduling apps can support restocking logistics, but they offer limited direct assistance to the core cleaning task itself.
Task automatability	claude-haiku-4-5-20251001	2/5	Limited automation is technically possible for supply restocking (inventory tracking, ordering), but the core physical tasks of cleaning (toilets, floors, mirrors) require dexterous manipulation and real-time adaptation to varying dirt, clogs, and spatial constraints that current robotics cannot reliably perform end-to-end. No current system achieves 50% time savings across the full task.
Task automatability	claude-sonnet-5	1/5	Physical restroom cleaning requires manipulation, judgment about mess types, and fine motor skills that current robots/AI cannot perform reliably or generally today.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While no license is required to automate restroom cleaning, adoption faces practical barriers including facility layout variability, need for human oversight of chemical safety and quality assurance, and customer/worker preference for human presence and accountability in hygiene-critical spaces. These are organizational rather than legal barriers.
Adoption barriers	claude-sonnet-5	2/5	No licensing barrier, but practical barriers exist: physical environments vary, hygiene liability, and customer expectations of human oversight in shared restrooms.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Autonomous cleaning systems (robots, chemical dispensers, IoT sensors) remain significantly more expensive to acquire, maintain, and integrate than the loaded wage of a janitor, even where partial automation exists. The breakeven point is distant for this low-skill, labor-intensive task.
Cost vs. human wage	claude-sonnet-5	1/5	Specialized cleaning robots remain expensive, need human supervision/restocking, and cannot match a janitor's flexibility and cost-effectiveness for this task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product today reliably performs full restroom cleaning autonomously in production environments. Specialized cleaning robots exist only in narrow research or pilot contexts and cannot handle the diversity of restroom layouts, soiling patterns, and edge cases encountered in real facilities.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs full restroom servicing autonomously; existing cleaning robots handle only narrow subtasks like floor mopping in controlled environments.

Clean windows, glass partitions, or mirrors, using soapy water or other cleaners, sponges, or squeegees.

CI 15–24 · exposure 8 · augmentation 13 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Janitor and cleaner roles remain in low-digitization, physical-work sectors with limited AI adoption patterns. Window cleaning is a routine manual task in small and medium facilities where cost-benefit of automation is unfavorable and adoption remains negligible.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial services are a low-digitization, physically-demanding sector with minimal AI/robotics adoption for routine cleaning tasks like glass cleaning.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI tools offer minimal assistance to a human cleaner performing this task; while scheduling or water-quality monitoring tools could exist, they do not meaningfully transform the core manual labor of physically cleaning glass surfaces.
Augmentation potential	claude-sonnet-5	1/5	Current AI tools offer essentially no assistance to a human performing manual window or mirror cleaning.
Task automatability	claude-haiku-4-5-20251001	2/5	While some window-cleaning robots exist in research and limited deployment, current AI systems cannot reliably perform end-to-end window cleaning across varied surfaces, angles, and environments with the dexterity and physical manipulation required. The task involves real-world contact and judgment that fall short of the 50% time-saving threshold for automated systems in production.
Task automatability	claude-sonnet-5	1/5	This is a physical dexterity task requiring manipulation of cleaning tools across varied surfaces and heights; no off-the-shelf AI system can perform it end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	2/5	The task occurs in physical spaces with safety concerns (heights, ladders, equipment operation) and occupational regulations, creating some friction for automation, though no hard legal requirement that a licensed human must perform it exists.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human, but practical barriers include equipment cost, need for adaptability to varied surfaces, and liability for damage/breakage.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current robotic window-cleaning systems cost thousands to tens of thousands of dollars with high maintenance, integration, and operational costs, vastly exceeding the cost of a human janitor performing the same work.
Cost vs. human wage	claude-sonnet-5	1/5	Specialized robotic cleaning equipment is far more expensive to acquire, deploy, and maintain than paying a janitor's hourly wage for this simple task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No mainstream commercial product reliably performs this task autonomously today. Experimental robotic cleaners exist but are narrow in scope, unreliable on varied glass types, and require significant setup and human oversight to function in real buildings.
Technical feasibility today	claude-sonnet-5	1/5	Window-cleaning robots exist only in narrow research/niche commercial contexts (e.g., high-rise facade robots at pilot scale) and are not deployed broadly for general janitorial glass cleaning.

Clean and polish furniture and fixtures.

CI 15–24 · exposure 8 · augmentation 13 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Janitor work is concentrated in small to mid-sized organizations with limited automation infrastructure; adoption of advanced cleaning robots remains nearly nonexistent in the sector.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and cleaning services are a low-digitization, physical-labor sector with minimal AI/robotics adoption for detailed cleaning tasks like furniture polishing.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Current AI offers minimal assistance—basic task scheduling or inventory management for supplies could help, but no AI system meaningfully augments the hands-on cleaning and polishing work itself.
Augmentation potential	claude-sonnet-5	1/5	Current AI tools offer negligible assistance for the physical act of cleaning and polishing furniture; there's no meaningful software-based productivity boost for this manual task.
Task automatability	claude-haiku-4-5-20251001	2/5	While some robotic systems exist for floor cleaning, polishing furniture and fixtures requires dexterous manipulation, real-time adaptation to varied surfaces and finishes, and quality judgment that current AI/robotics cannot reliably perform end-to-end with 50% time savings at equal quality.
Task automatability	claude-sonnet-5	1/5	Cleaning and polishing furniture requires physical dexterity, mobility, and adaptability to varied surfaces/objects that current robotics cannot handle at general-purpose scale; no off-the-shelf AI system performs this physical task end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	2/5	There are minimal regulatory or licensing barriers to automation, but customer preferences for human inspection of high-value fixtures and the physical damage risk from failed automation attempts create moderate organizational friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human for this task, but physical environment variability and lack of technology create practical (not regulatory) barriers to substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Robotics capable of this task (custom manipulators, vision systems, specialized end-effectors) cost tens of thousands of dollars with ongoing maintenance, far exceeding the loaded wage cost of a janitor performing the task.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI/robotic solution for this task, so any hypothetical automation would require expensive custom robotics far exceeding the cost of human labor for this task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial product reliably cleans and polishes furniture and fixtures autonomously in real work environments; prototype robotics exist but require heavy human supervision and intervention.
Technical feasibility today	claude-sonnet-5	1/5	No deployed consumer or commercial product autonomously cleans and polishes diverse furniture and fixtures; existing robotic cleaners (e.g., floor vacuums) address a narrow adjacent task, not this one.

Monitor building security and safety by performing tasks such as locking doors after operating hours or checking electrical appliance use to ensure that hazards are not created.

CI 5–33 · exposure 13 · augmentation 38 · 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	Adoption of AI for security monitoring in institutional facilities (schools, offices, factories) remains limited to large enterprises; small and mid-sized buildings—the majority of janitor employment—lack investment in automation and rely on human security rounds.
Sector adoption velocity	claude-sonnet-5	1/5	Building services and janitorial work are low-digitization, physical-labor sectors with minimal AI/robotics adoption for these specific security-check tasks.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-powered security cameras with alert systems can assist janitors by flagging potential hazards or unusual door-lock states, reducing manual inspection time; however, the human must still validate and act on alerts, limiting augmentation to partial workflow support.
Augmentation potential	claude-sonnet-5	2/5	Smart locks, motion sensors, and IoT appliance monitors can alert staff to issues, offering modest assistance, but the core walk-through and judgment task remains largely unaided by AI.
Task automatability	claude-haiku-4-5-20251001	2/5	Monitoring building security and safety requires physical presence, judgment about hazard detection, and response to contextual variations. While AI could assist with some monitoring (e.g., via video analysis), the task inherently demands human judgment about electrical safety, irregular hazards, and the authority to lock doors—current systems cannot reliably automate the full scope end-to-end.
Task automatability	claude-sonnet-5	1/5	This requires physical presence, mobility, and situational judgment to physically inspect a building, lock doors, and check appliances—no current AI system can perform these physical actions end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Building security and safety are heavily regulated; liability for missed hazards or unauthorized door locking creates substantial legal exposure; many jurisdictions require human accountability for security decisions, and organizations retain strong preference for human presence during operating hours.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement for this specific task, but physical liability for security lapses and property damage creates some organizational caution around removing human oversight.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The infrastructure cost for security cameras, access control systems, and integration with building management, plus continuous monitoring oversight, compares unfavorably to the wage of a single custodian performing these tasks during shifts.
Cost vs. human wage	claude-sonnet-5	1/5	Achieving equivalent physical security coverage would require robotics and IoT infrastructure investment that far exceeds the wage cost of a janitor performing this task.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Security camera monitoring systems with basic anomaly detection exist, but deploying AI to autonomously lock doors and make safety judgments about electrical hazards lacks reliable real-world production systems at scale; most deployments remain pilot-stage with significant oversight requirements.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously patrols buildings, locks doors, and inspects appliance hazards; some smart-building sensors exist but they don't replace the physical monitoring task itself.

Mow or trim lawns or shrubbery, using mowers or hand or power trimmers, and clear debris from grounds.

CI 5–33 · exposure 13 · augmentation 25 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Janitors and grounds maintenance remain low-digitization, small-firm-dominated sectors. Adoption of robotic mowers is still pilot-stage and limited to large institutional or wealthy residential settings; mainstream commercial grounds maintenance continues to rely on manual labor.
Sector adoption velocity	claude-sonnet-5	1/5	Grounds maintenance and janitorial services are a physical, low-digitization sector with minimal AI/robotics adoption in production settings.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Robotic mowers and power trimmers offer modest productivity gains by reducing manual repetition, but they do not materially augment human capability to handle complex debris removal, shrubbery design decisions, or variable site conditions.
Augmentation potential	claude-sonnet-5	2/5	Power tools already assist manual labor but AI specifically offers little additional productivity boost for this outdoor physical task beyond existing mechanized equipment.
Task automatability	claude-haiku-4-5-20251001	2/5	While lawn-mowing robots exist and can handle simple, regular lawns, they struggle with complex grounds, obstacles, debris removal, and trimming varied shrubbery. The full task—including debris clearing and varied trimming—requires significant human judgment and manual dexterity that current systems cannot reliably perform end-to-end.
Task automatability	claude-sonnet-5	1/5	Robotic mowers exist for simple residential lawns but this task encompasses varied grounds, trimming shrubbery, and debris clearing across irregular commercial/institutional properties, which current autonomous systems cannot handle end-to-end at equal quality.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Safety liability for autonomous equipment on commercial grounds, property damage risk, and neighborhood/HOA restrictions on robotic lawn care create meaningful adoption friction; human operators are often preferred for liability and quality control reasons.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but liability concerns (property damage, safety around obstacles/people) and organizational reliance on flexible human labor create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Autonomous mowers are capital-intensive ($1,000–$5,000+) with maintenance costs; total cost of ownership per task execution remains higher than or comparable to hiring a laborer, especially for variable, non-routine grounds.
Cost vs. human wage	claude-sonnet-5	1/5	Robotic mowing equipment plus setup, maintenance, and supervision costs typically exceed or match human labor costs for this multi-task, variable-terrain job, especially given need for a human to handle trimming and debris removal.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Autonomous mowing robots are deployed in limited, controlled settings (residential lawns, golf courses), but they cannot reliably handle debris removal, obstacle navigation, or shrubbery trimming at production quality without human setup and oversight.
Technical feasibility today	claude-sonnet-5	1/5	Consumer robotic mowers are deployed but limited to flat, enclosed residential lawns; no product reliably trims shrubbery or clears general debris in production at scale for janitorial contexts.

Follow procedures for the use of chemical cleaners and power equipment to prevent damage to floors and fixtures.

CI 15–21 · exposure 0 · augmentation 25 · 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	2/5	Cleaning is a laggard sector with high fragmentation across small facilities and property management firms; adoption of automation is slow and concentrated in large commercial real estate and hospitality chains, not representative of the occupation broadly.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial services are a low-digitization, physical-labor sector with minimal AI/robotic adoption in production beyond niche autonomous floor scrubbers.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI offers minimal augmentation for this task—perhaps scheduling optimization or inventory tracking of chemicals, but not for the core judgment of procedure selection and real-time equipment operation where human skill and responsiveness remain essential.
Augmentation potential	claude-sonnet-5	2/5	AI could provide reference guidance or training materials on chemical procedures, but offers little real-time assistance during the physical task itself.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires real-time visual inspection, spatial reasoning, and adaptive physical handling of equipment to prevent damage—capabilities that current AI systems cannot perform end-to-end in unstructured physical environments. The judgment of when and how to apply cleaners and power equipment to different materials remains beyond autonomous systems today.
Task automatability	claude-sonnet-5	1/5	This requires physical dexterity, judgment about surface conditions, and manual operation of cleaning equipment in real-world spaces, which current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Liability for equipment damage and chemical safety mishaps creates moderate friction; OSHA regulations govern chemical handling but do not explicitly mandate human performance. Property damage from incorrect equipment use and organizational preference for human accountability provide some protection against full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but liability for chemical misuse and property damage creates some caution around unsupervised automation, plus physical environment variability is a practical barrier.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Autonomous cleaning robots and chemical dispensers exist but are expensive to purchase, maintain, and integrate; their cost per cleaned area typically exceeds hiring janitors at minimum wage in most markets, especially when factoring in supervision and failure recovery.
Cost vs. human wage	claude-sonnet-5	1/5	Human janitorial labor remains far cheaper than any robotic/AI system capable of this physical, judgment-based task, given equipment costs and limited capability.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product can autonomously follow chemical safety protocols and operate power equipment on varied floor types while preventing damage in real buildings. While cleaning robots exist, they operate in narrow, pre-mapped environments and lack the adaptive chemical knowledge and damage-prevention judgment this task demands.
Technical feasibility today	claude-sonnet-5	1/5	There are no deployed AI/robotic products that reliably follow chemical safety procedures and operate power cleaning equipment across varied real-world settings; robotic floor cleaners exist but do not handle this full scope.

Steam-clean or shampoo carpets.

CI 10–24 · exposure 8 · augmentation 25 · importance 3.0/5 · 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	Janitorial services are typically low-digitization, small-firm dominated sectors with minimal AI/robotic adoption; even large facilities cleaning budgets lag far behind information-sector automation uptake.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and building services is a low-digitization, physically-oriented sector with minimal AI/robotics adoption for deep cleaning tasks specifically.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could marginally assist by monitoring carpet condition or scheduling cleaning routes, but the core physical task of steam-cleaning or shampooing offers limited room for human-in-the-loop AI augmentation.
Augmentation potential	claude-sonnet-5	2/5	Some smart equipment (sensors, scheduling apps) can help plan or track cleaning tasks, but current AI offers little direct assistance to the physical act of steam-cleaning or shampooing carpets.
Task automatability	claude-haiku-4-5-20251001	2/5	Steam-cleaning and shampooing carpets require physical manipulation of equipment, positioning in tight spaces, and assessing carpet condition—capabilities current AI robots lack at production scale. While some robotic floor-cleaning systems exist, they cannot reliably handle the full task end-to-end (setup, application, drying assessment, edge work).
Task automatability	claude-sonnet-5	1/5	This is a physical manipulation task requiring equipment handling, mobility, and dexterity across varied surfaces; no off-the-shelf AI system performs this end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	3/5	No strict licensing requirement for carpet cleaning itself (in most jurisdictions), but organizational reliance on skilled manual labor, customer expectations for human attention to quality, and the capital outlay for reliable automation create moderate friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal barrier prevents automation, but physical equipment operation, liability for equipment damage, and access to varied real-world spaces create practical friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized carpet cleaning equipment and robotic systems capable of this work are significantly more expensive to acquire, operate, and maintain than paying a janitor to perform it manually.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI/robotic substitute at scale, so the human laborer remains the only cost-effective option; any experimental robotic solution would be far more expensive than a janitor's wage.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial product reliably performs full carpet steam-cleaning or shampooing autonomously at scale in real-world facilities today. Robotic floor-cleaning exists in narrow contexts, but carpet-specific deep cleaning remains firmly in the manual domain.
Technical feasibility today	claude-sonnet-5	1/5	No deployed commercial product performs autonomous steam-cleaning or shampooing of carpets; robotic vacuums exist but this specific deep-cleaning task is not addressed by mature robotics products.

Move heavy furniture, equipment, or supplies, either manually or with hand trucks.

CI 15–15 · 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	Adoption of automation for this task is minimal even in large facilities. Janitorial work remains highly manual and labor-reliant; facility managers have not widely invested in robotics for general moving tasks due to cost, reliability, and flexibility constraints.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and physical labor sectors show minimal AI/robotics adoption for manual material handling, lagging far behind information-work automation trends.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI offers minimal augmentation for physical moving tasks. Exoskeletons could theoretically assist, but are not yet mainstream in janitorial settings; current AI assistants cannot meaningfully improve the core physical labor of manual furniture relocation.
Augmentation potential	claude-sonnet-5	2/5	Hand trucks and basic mechanical aids already assist this task, but AI specifically adds little beyond existing simple tools; no meaningful AI-driven productivity gain.
Task automatability	claude-haiku-4-5-20251001	1/5	Current AI systems lack the embodied physical capability to move heavy furniture, equipment, or supplies either manually or with mechanical aids. This task requires dexterous manipulation, spatial reasoning in real environments, and significant force application—capabilities that do not exist in deployed robotic systems at consumer or facility scales.
Task automatability	claude-sonnet-5	1/5	Physical manipulation of heavy furniture and equipment requires mobile robotics with dexterity and strength far beyond current off-the-shelf systems; no general AI can perform this end-to-end today.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	While there are no strict legal licensing barriers to automation, workplace safety liability, customer expectations of human workers, and the need for judgment about damage risk to facilities and contents create moderate friction against substitution.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but physical safety concerns, liability for damage/injury, and building/space variability create moderate practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current robotic systems capable of any furniture moving are prohibitively expensive (six figures+) compared to hourly wages for manual labor, and require substantial infrastructure and maintenance, making all-in costs far higher than human workers.
Cost vs. human wage	claude-sonnet-5	1/5	Specialized robotic hardware for heavy object transport is far more expensive than a human worker for this task, with no viable off-the-shelf substitute.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No mainstream deployed product reliably performs general furniture and equipment moving in diverse facility settings. Specialized industrial robots exist for narrow, controlled environments (warehouses with pallets), but not for the varied, unstructured indoor spaces where janitors work.
Technical feasibility today	claude-sonnet-5	1/5	No deployed commercial product reliably moves heavy furniture/equipment in varied janitorial settings; existing warehouse robots are narrow, structured, and not applicable to general moving tasks.

Set up, arrange, or remove decorations, tables, chairs, ladders, or scaffolding to prepare facilities for events, such as banquets or meetings.

CI 15–15 · exposure 0 · augmentation 13 · 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	Adoption remains negligible; facility and event management sectors show little production-level automation of setup tasks, relying instead on manual labor due to the physical and spatial complexity involved.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and facilities services are a low-digitization, physical-labor sector with minimal AI/robotics adoption for this kind of task.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI offers minimal meaningful assistance in this task. Planning tools or floor-layout software could theoretically help with arrangement visualization, but the core physical execution remains entirely human-dependent without significant augmentation benefit.
Augmentation potential	claude-sonnet-5	1/5	Current AI tools (e.g., scheduling apps) offer negligible assistance to the core physical act of arranging furniture and equipment.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical manipulation of heavy objects, spatial arrangement based on event-specific layouts, and navigation of real-world constraints. Current AI systems cannot physically move furniture or execute real-world setup tasks without specialized robotics that remain highly constrained in general environments.
Task automatability	claude-sonnet-5	1/5	This is a physical manipulation task involving furniture, decorations, and heavy equipment across varied layouts, which current AI systems cannot perform end-to-end; robotics for this remains research-stage.'
Adoption barriers	claude-haiku-4-5-20251001	2/5	There are minimal legal or regulatory barriers to automating this task, though liability concerns around property damage and the need for human judgment on aesthetic arrangement create modest friction to full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human specifically, but physical safety concerns around ladders/scaffolding and lack of technology create practical friction rather than hard barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized robotic systems capable of furniture moving and spatial arrangement are extremely expensive to deploy, integrate, and maintain, making them far more costly than hiring a janitor for event setup tasks.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI/robotic substitute at any cost for this manual labor task, so humans remain far cheaper than any hypothetical automated solution.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial products reliably perform this task end-to-end. Robotic manipulation systems exist in research and limited industrial settings, but none operate autonomously at the speed and flexibility needed for general facility setup across diverse event types.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs general-purpose event setup/teardown involving tables, chairs, ladders, and scaffolding in unstructured environments today.

Drive vans, industrial trucks, or other vehicles required to travel to, or to perform, cleaning work.

CI 0–19 · exposure 8 · 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	Cleaning services remain highly fragmented, employ small firms with low capital budgets, and operate in physical, low-digitization environments. Adoption of autonomous fleets for this work is near zero in real organizations; the sector is a laggard in vehicle automation.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and facilities services are a low-digitization, physical-labor sector with minimal AI/robotics adoption for combined driving and cleaning duties.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI-assisted route planning, scheduling, and fuel optimization exist and help janitor fleet managers, but on-board vehicle automation (autonomous driving) offers minimal augmentation to the driver themselves. Most benefit is administrative rather than task-level augmentation.
Augmentation potential	claude-sonnet-5	2/5	AI could assist with route planning or scheduling for travel between cleaning sites, but offers little to no assistance for the physical driving or cleaning itself.
Task automatability	claude-haiku-4-5-20251001	2/5	While autonomous vehicle technology exists, it is not yet widely deployed for general cleaning route work. Current AI can navigate structured routes but struggles with the unpredictable environments, parking logistics, and real-time decision-making required for mobile cleaning operations. The task is partially automatable in theory but not at the 50% time-saving threshold with existing commercial systems.
Task automatability	claude-sonnet-5	1/5	Driving vehicles combined with variable cleaning tasks requires physical manipulation and navigation in unstructured environments that current AI systems cannot handle end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Autonomous vehicle operation faces hard regulatory barriers (licensing, liability frameworks, safety certification) and requires human oversight or remote operation. Most jurisdictions require a licensed human operator or supervisor in or monitoring the vehicle, creating a legal requirement that blocks full automation.
Adoption barriers	claude-sonnet-5	3/5	Driving vehicles on public roads requires licensing and liability coverage, adding regulatory friction beyond typical cleaning tasks, though not as strict as professional licensure fields.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current autonomous vehicle solutions are expensive per hour (leasing, maintenance, liability insurance, remote operators as fallback) and far exceed the wage cost of a janitor driver. Integration and oversight overhead add further cost.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI system for this combined task, so AI cost is effectively infinite relative to a human driver-cleaner's wage.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed, production-scale autonomous vehicle service exists for routine cleaning fleet logistics. Autonomous vehicles are in limited pilots in controlled environments; they do not reliably operate across the diverse, unstructured environments where janitors must travel and park.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product combines autonomous vehicle operation with general cleaning work; this remains far beyond current commercial robotics or AI systems.

Clean chimneys, flues, and connecting pipes, using power or hand tools.

CI 5–10 · exposure 0 · augmentation 13 · importance 3.2/5 · 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	The chimney-cleaning sector remains highly fragmented, locally operated, and low-digitization. Adoption of specialized automation is minimal; most work is still performed by traditional trade workers using hand and power tools.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and facilities maintenance work is a low-digitization, physical-labor sector with minimal AI/robotic adoption for tasks like this.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Current tools (cameras, vacuum systems, inspection drones) can assist a human chimney sweep with diagnostics and debris removal, but the core task of navigating confined spaces and ensuring thorough cleaning remains largely manual and human-dependent.
Augmentation potential	claude-sonnet-5	1/5	Current AI offers essentially no meaningful assistance for the physical act of cleaning chimneys, flues, and pipes using tools.
Task automatability	claude-haiku-4-5-20251001	1/5	Chimney cleaning requires physical navigation of confined spaces, tactile feedback to assess debris buildup, and real-time hazard detection in variable conditions. Current robots cannot reliably perform end-to-end chimney cleaning across diverse chimney types and configurations at 50% time savings today.
Task automatability	claude-sonnet-5	1/5	This is a physical manual task requiring dexterity, tool manipulation, and navigating confined spaces like chimneys and flues; no AI system can perform this end-to-end today.rapper
Adoption barriers	claude-haiku-4-5-20251001	4/5	Chimney inspection and cleaning often require licensed chimney sweeps or certified technicians in many jurisdictions. Liability for structural damage or fire hazard creates strong incentive for human expertise and certification, forming a substantial regulatory and legal barrier.
Adoption barriers	claude-sonnet-5	3/5	No licensing typically required for basic janitorial chimney cleaning, but physical access constraints, safety concerns, and specialized tool handling create moderate practical barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Specialized chimney-cleaning robots are expensive capital equipment with high maintenance costs, while human chimney sweeps command modest wages. The cost per task remains well above human labor, particularly for the long tail of non-standard chimneys.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI or robotic substitute, so any AI-based approach would be more expensive than simply hiring a human cleaner.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	While some specialized chimney-cleaning robots exist in research or limited commercial settings, they are narrow-domain, unreliable on complex blockages, and require significant human oversight. No deployed product reliably performs full chimney cleaning autonomously at production scale.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product or robot performs chimney/flue cleaning in production; this remains a purely manual trade task.

Make adjustments or minor repairs to heating, cooling, ventilating, plumbing, or electrical systems.

CI 0–14 · exposure 8 · augmentation 25 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Janitorial and facility-maintenance sectors are historically low-tech, composed largely of small operators and budget-constrained institutions. Adoption of advanced diagnostics or repair automation remains minimal; most facilities rely on on-call licensed contractors rather than in-house AI systems.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and facilities maintenance is a low-digitization, physically-embodied sector with minimal AI/robotics adoption for repair tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI can provide diagnostic guidance (e.g., flowcharts for basic troubleshooting) that assists a technician, but current tools offer only marginal productivity gains on the core repair task itself. The hands-on nature of the work limits how much AI presence meaningfully augments the human worker's output.
Augmentation potential	claude-sonnet-5	2/5	AI can help via diagnostic apps, manuals, or troubleshooting guidance, but offers limited assistance for the hands-on repair work itself.
Task automatability	claude-haiku-4-5-20251001	2/5	This task requires physical manipulation of building systems, spatial reasoning, and real-time problem diagnosis in complex environments. While AI can assist with diagnostic logic (e.g., troubleshooting flowcharts), the hands-on repair work—accessing confined spaces, turning valves, replacing components—cannot be performed by current AI without specialized robotics, which remain rare in janitorial contexts.
Task automatability	claude-sonnet-5	1/5	This requires physical manipulation, diagnosis, and manual repair of building systems in varied unstructured environments—no current AI system can perform this end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Electrical and plumbing work in most jurisdictions requires licensed or certified technicians; liability exposure for faulty repairs is high, creating legal and insurance barriers. Building codes and safety regulations mandate human sign-off, creating hard adoption barriers regardless of technical capability.
Adoption barriers	claude-sonnet-5	4/5	Electrical and plumbing work often falls under licensing, safety codes, and liability requirements, and physical access/repair inherently requires human presence.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current AI diagnostic tools cost comparable to or more than direct human labor when integrated into real workflows. The physical robotics needed to match human hands-on repair capability remain capital-intensive and uneconomical versus hiring skilled or semi-skilled technicians.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-only substitute for physical repair labor, so AI cost comparison is not applicable and the human remains far cheaper than any hypothetical robotic solution today.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed products reliably perform HVAC, plumbing, or electrical repairs end-to-end in production janitorial settings. Diagnostic AI tools exist but require human technicians to execute repairs; autonomous repair robots for building systems are still research/prototype stage.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs physical HVAC/plumbing/electrical repairs autonomously; robotics for general building maintenance remains research-stage.

Spray insecticides or fumigants to prevent insect or rodent infestation.

CI 5–5 · exposure 0 · augmentation 13 · 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	Janitorial and facility management is a laggard sector with low digitization, small-firm dominance, and physical on-site requirements. Adoption of any automation in this task remains negligible; most facilities still rely on trained human staff for pest control.
Sector adoption velocity	claude-sonnet-5	1/5	Janitorial and pest control work is a low-digitization, physically intensive sector with minimal AI/robotic adoption for chemical application tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with scheduling spraying cycles or identifying infestation hotspots via image analysis, but the core spraying activity itself involves minimal opportunity for AI assistance; the task is fundamentally a physical execution step with regulatory compliance as the primary cognitive load.
Augmentation potential	claude-sonnet-5	1/5	AI offers little to no meaningful assistance for the physical act of spraying chemicals, though scheduling or record-keeping software might tangentially help but not the core task.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires physical manipulation of spraying equipment in diverse indoor and outdoor environments, navigation around obstacles, and real-time judgment about infestation severity and chemical safety. Current robotics cannot reliably perform end-to-end pesticide application with the dexterity, mobility, and environmental adaptability this task demands.
Task automatability	claude-sonnet-5	1/5	This is a physical task requiring navigating a building, identifying infestation-prone areas, and manually applying chemicals; no AI system can perform the physical spraying or judgment-based site assessment today.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Pesticide application is regulated by EPA and local codes; human workers must be licensed or certified to apply certain chemicals, and liability for improper application (health damage, environmental harm) creates legal responsibility that must remain with a qualified human. These licensing and liability barriers substantially protect the task.
Adoption barriers	claude-sonnet-5	4/5	Applying insecticides/fumigants often requires certification or licensing depending on jurisdiction, and mishandling creates significant liability and safety risk, creating meaningful regulatory and legal barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The hardware cost for a robot capable of safe autonomous pesticide spraying, plus regulatory compliance, oversight, and integration, far exceeds the loaded wage of a janitor performing this task, which is typically low-cost labor.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI-driven substitute for physical chemical application, so AI cost is not comparable—human labor (or specialized pest control robots, which are not AI in the relevant sense and are costly) remains necessary.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed commercial products today perform autonomous pesticide spraying for general janitorial use. Specialized agricultural drones exist but are not integrated into building maintenance workflows, and pesticide application remains a human-performed task even in highly digitized facilities.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs pest control spraying autonomously; this remains a manual task performed by licensed applicators or janitorial staff.

Related occupations — Building & Grounds Cleaning & Maintenance

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.