Recycling and Reclamation Workers
53-7062.04Median wage $40,240/yr2,950,280 employed (US)Rank #446 of 923 scored · top 48% by substitution
Prepare and sort materials or products for recycling. Identify and remove hazardous substances. Dismantle components of products such as appliances.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution29
Exposure21
Augmentation32
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%21
panel mean rating 1.9/5 → substitution pressure 21/100
Technical feasibility todayw 20%21
panel mean rating 1.8/5 → substitution pressure 21/100
Cost vs. human wagew 15%21
panel mean rating 1.8/5 → substitution pressure 21/100
Adoption barriersw 20%inverted — strong barriers lower the score62
panel mean rating 2.5/5 (barrier strength) → substitution pressure 62/100
Sector adoption velocityw 10%15
panel mean rating 1.6/5 → substitution pressure 15/100
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 logs of recycled materials or waste chemicals removed from products.
73CI 70–76 · exposure 75 · augmentation 75 · importance 3.9/5 · click for rater detail
Record logs of recycled materials or waste chemicals removed from products.
73| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Recycling facilities are adopting automated logging incrementally, with large municipal and industrial operators leading, but many small operations still rely on manual tracking. Adoption is uneven and varies by facility funding and digitization maturity. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Waste and recycling is a physically-oriented, less digitized sector where adoption of automated tracking systems is growing but still uneven, especially among smaller reclamation operations. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI assistance dramatically improves worker productivity by auto-capturing material data via cameras or sensors, reducing manual data-entry errors and freeing workers for higher-value sorting and quality tasks. Workers remain involved but benefit from significant cognitive offloading. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enabled scanning, OCR, and data systems significantly speed up and reduce errors in logging tasks while workers still handle physical sorting and verification. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can reliably extract, categorize, and log material composition data from product barcodes, images, or chemical databases with minimal human intervention. The task is primarily data entry and classification—highly suitable for automation—though some edge cases (ambiguous materials, non-standard products) may require human review, preventing a full 5. |
| Task automatability | claude-sonnet-5 | 4/5 | Recording logs of quantities and types of recycled materials is a structured data-entry task that can largely be automated via barcode/RFID scanning, sensors, and software integration with minimal human input. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No legal licensing requirement or human sign-off mandate exists for logging recycled materials; the main friction is integration with existing facility workflows and staff acceptance. Barriers are low relative to regulated professions. |
| Adoption barriers | claude-sonnet-5 | 2/5 | Some regulatory recordkeeping requirements exist for hazardous waste chemicals, but the act of logging itself isn't restricted to a licensed human and can be system-generated with review. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated logging via vision/scanning and database entry costs cents per item once infrastructure is amortized, whereas manual record-keeping requires paid labor for each entry. The cost disparity is at least an order of magnitude in favor of AI systems. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated scanning and digital logging systems are far cheaper per transaction than manual record-keeping once installed, though initial hardware/software integration adds cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Computer vision systems, barcode/RFID scanners, and automated logging platforms are deployed in recycling facilities today for material tracking and inventory management. Production systems exist and operate at scale, though accuracy varies by material type and facility sophistication, preventing a full 5. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Inventory and waste-tracking software with scanning and automated logging is already deployed in many recycling and waste management facilities, though smaller operations still use manual logs. |
Sort metals to separate high-grade metals, such as copper, brass, and aluminum, for recycling.
53CI 35–71 · exposure 42 · augmentation 50 · importance 4.0/5 · click for rater detail
Sort metals to separate high-grade metals, such as copper, brass, and aluminum, for recycling.
53| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Metal recycling is a capital-intensive, digitized industry where large operators (Sims Metal, Alegheny Technologies, etc.) are rapidly deploying automated sorting. Measurable displacement of manual sorting is documented in developed-market recycling centers, with vendors reporting steady uptake across North America and Europe. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | The waste and recycling industry is a physical, lower-digitization sector where automation adoption is slow and concentrated in large-scale materials recovery facilities rather than widespread across the sector. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted sorting tools can highlight uncertain metals for human verification or provide real-time composition feedback, improving decision-making on borderline alloys. However, augmentation is secondary to the primary automation use case in this task; workers largely monitor rather than actively sort when systems are operational. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-based sensor sorting and computer vision systems can assist human sorters by pre-sorting or flagging materials, improving speed and accuracy, though humans remain central to quality control and handling exceptions. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Automated sorting of metals is partially feasible using optical and electromagnetic sensors (X-ray fluorescence, eddy current) to identify material composition, but real-world contamination, mixed alloys, and irregular shapes create variability that typically requires human judgment or multi-stage processing. Current systems handle ~50% of sorting throughput with acceptable purity, meeting the partial automation threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Sorting metals by grade requires physical manipulation, visual/spectroscopic identification, and handling of varied irregular scrap, which is only partially automatable with current robotics and sensor systems, not a pure software task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | No licensing requirement or legal mandate to use human sorters; sorting is treated as an operational choice. Adoption is primarily driven by economics and facility automation. Minimal regulatory or liability barriers prevent substitution, though some operators prefer humans for certain specialty alloys. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or legal requirement mandates a human sorter; the barrier is mainly capital cost and material variability rather than regulatory or liability constraints. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Automated metal-sorting equipment costs tens of thousands of dollars upfront but processes tonnage continuously at per-unit costs well below human hand-sorting wages when annualized over high throughput; the payback period is typically 2–4 years in commercial recycling, making AI sorting substantially cheaper per sortable item. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Advanced sorting equipment (sensor-based sorters, robotic arms) requires significant capital investment, making it costlier than manual labor for many small-to-mid-size recycling operations despite long-term efficiency gains at scale. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed metal-sorting systems using AI-enhanced spectroscopy and sensor fusion exist in recycling facilities today; Tomra, STEINERT, and similar vendors operate production systems at scale. Minor gaps remain in edge cases (heavily corroded or coated metals), but core functionality is proven and operational. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated sorting systems (eddy current separators, XRF/LIBS sorters, AI vision-guided robotic pickers) exist in some advanced recycling facilities, but most reclamation work still relies heavily on manual sorting due to cost and variability of scrap streams. |
Operate processing equipment, such as fiber-sorters and grinders, to sort, crush, or grind recyclable materials.
37CI 30–44 · exposure 33 · augmentation 50 · importance 3.8/5 · click for rater detail
Operate processing equipment, such as fiber-sorters and grinders, to sort, crush, or grind recyclable materials.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Recycling facilities are capital-constrained, fragmented, and slower to digitize than information or finance sectors; pilot projects exist but widespread production adoption of AI-driven sorting remains limited and geographically uneven. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Waste and recycling is a low-digitization, physically intensive sector where automation adoption is slow and capital-constrained compared to information/professional sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted sorting systems can help workers identify contamination or material types, improving throughput and safety, but the task remains heavily equipment-operator-dependent and worker oversight is still required for quality control. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled optical sorting and predictive maintenance can assist workers in identifying materials and optimizing equipment use, improving efficiency while humans remain operationally involved. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Material sorting and grinding can be partially automated with existing computer vision and robotic systems, but current technology struggles with contamination detection, material variability, and safe equipment operation across diverse waste streams. Setup and oversight remain substantial, placing this near the half-automation threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical machine operation involving material handling, monitoring, and adjustment is not something current AI systems can perform end-to-end; automation here requires robotics/mechanization rather than AI software. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Workplace safety regulations and equipment operation licensing create some friction; however, no strict legal barrier prevents AI automation of sorting and grinding if safety standards are met. Organizational inertia and union concerns add moderate resistance. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but safety regulations around heavy machinery operation and physical workplace hazards create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Robotic sorters and AI vision systems remain capital-intensive and require ongoing maintenance, training, and integration costs that often exceed the loaded wage of recycling workers, especially in smaller or lower-throughput facilities. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Advanced sorting equipment with AI vision systems is capital-intensive and often costs more than human labor for small-to-mid scale operations, though large facilities may see gains. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While robotic sorting pilots exist in recycling facilities, production deployment remains limited; most systems handle narrow material categories, have error rates requiring human intervention, and face integration challenges with legacy processing equipment. No mature end-to-end automation products are standard across the industry. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some automated sorting systems (optical/AI-based sorters) exist in modern facilities, but widespread reliable operation of grinders and fiber-sorters by AI without human operators is not standard in production. |
Sort materials, such as metals, glass, wood, paper or plastics, into appropriate containers for recycling.
35CI 35–35 · exposure 25 · augmentation 38 · importance 4.3/5 · click for rater detail
Sort materials, such as metals, glass, wood, paper or plastics, into appropriate containers for recycling.
35| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption is concentrated in large, well-capitalized recycling plants and advanced sorting facilities; small and mid-sized operations (where most sorting workers are) lag significantly. Sector digitization remains uneven, and capital constraints limit velocity in price-sensitive waste management. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Waste management and recycling is a low-digitization, physical-labor-heavy sector where robotic sorting adoption is growing but remains concentrated in a minority of large, well-funded facilities. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted vision systems can help workers identify borderline cases and flag contamination, improving accuracy and speed on portions of the task. However, the core sorting operation remains physically manual, so augmentation is meaningful but limited to decision support rather than transformative productivity gain. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI-assisted optical sorting can help human workers identify materials faster or reduce contamination, but this augmentation is limited to facilities with such technology already installed, not a general productivity boost for most workers. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Material sorting has high potential for automation in controlled environments, but current AI vision systems struggle with real-world contamination, opacity of mixed streams, and varied material states. Partial automation (pre-sorting lines, basic segregation) exists, but end-to-end replacement meeting 50% time savings at equal quality remains unreliable without significant infrastructure setup. |
| Task automatability | claude-sonnet-5 | 2/5 | While AI-powered optical sorting systems exist for industrial recycling facilities, the manual task of a human worker sorting mixed materials by hand remains largely unautomated in most contexts; robotic sorting requires substantial capital infrastructure not yet ubiquitous. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Regulatory and safety barriers are low (no licensing required for the sorting task itself), but operational friction is moderate: facility retrofitting, worker retraining needs, and equipment maintenance requirements slow adoption. Organizational inertia in fragmented, often small-scale waste management firms creates friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates human sorting, but physical infrastructure retrofitting, capital costs, and the messy/variable nature of waste streams create meaningful practical barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized robotic sorting systems are capital-intensive (hundreds of thousands of dollars) with high integration costs, making per-unit task cost comparable to or higher than human labor in low-to-mid-throughput settings. Only high-volume facilities begin to achieve cost parity. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Robotic sorting systems require significant upfront capital investment (robotic arms, sensors, conveyor integration) that only pays off at high-volume facilities; for smaller operations, human labor remains cheaper on a per-unit basis. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Robotic sorting and AI-guided picking systems exist in some recycling facilities, but error rates remain material (misidentification of contaminated or ambiguous items), and deployment is limited to high-throughput industrial plants. Most facilities still rely on manual labor; no mature off-the-shelf product reliably handles the full diversity of materials and contamination in production. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI/robotic sorting systems (e.g., AMP Robotics, optical sorters) are deployed in some large-scale material recovery facilities, but most recycling sorting is still done manually or with basic mechanical separation, not advanced AI-driven robotic picking at broad scale. |
Deposit recoverable materials into chutes or place materials on conveyor belts.
35CI 35–35 · exposure 25 · augmentation 25 · importance 4.2/5 · click for rater detail
Deposit recoverable materials into chutes or place materials on conveyor belts.
35| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Recycling facilities are often small operators with limited digitization and capital investment capacity. While some larger operations are piloting automation, the sector overall shows slow adoption of AI-driven material handling compared to information-intensive industries. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Waste and recycling is a physically-oriented, lower-digitization sector where robotic automation adoption is happening but slowly and unevenly compared to information-sector AI adoption. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist with real-time sorting guidance or line-speed optimization, but the core manual task of physical placement offers limited opportunity for meaningful AI augmentation while a human remains in the loop. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI-powered vision systems can assist by identifying material types for sorting, but this augments upstream sorting decisions more than the physical act of depositing materials into chutes. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires physical manipulation in a real-world setting with variable material types and positions. While robotic systems exist for sorting and material handling, end-to-end automation of depositing mixed recyclables into chutes or conveyor belts at the speed and quality of human workers remains technically challenging and not standard in deployed systems. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a physical manual sorting/handling task requiring mobility and dexterity in unstructured environments; current AI (software) cannot perform the physical act, though robotic sorting systems exist for narrow cases.dig |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory barriers legally require human labor for this task, but material variability, safety concerns, and the capital investment needed for automation create practical friction. Most facilities operate with human workers because it remains economically simpler. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal requirement mandates human performance, but physical workspace constraints, safety systems, and capital costs create moderate practical friction to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Robotic material-handling systems involve high capital costs and integration expenses that generally exceed the loaded wages of recycling workers, especially for the heterogeneous sorting and placement required in this task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Robotic sorting/conveyor systems require significant capital investment, integration, and maintenance, often exceeding the cost of low-wage manual labor for this task in most facilities. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Specialized robotic systems for material handling and sorting exist in some facilities, but they typically handle homogeneous streams and require significant setup. General-purpose automation that reliably deposits diverse recoverable materials into chutes across typical recycling operations is not yet mature in production. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Robotic sorting arms with AI vision are deployed in some material recovery facilities but are narrow, expensive installations, not general replacements for the manual placement task described. |
Operate forklifts, pallet jacks, power lifts, or front-end loaders to load bales, bundles, or other heavy items onto trucks for shipping to smelters or other recycled materials processing facilities.
25CI 23–28 · exposure 25 · augmentation 25 · importance 4.0/5 · click for rater detail
Operate forklifts, pallet jacks, power lifts, or front-end loaders to load bales, bundles, or other heavy items onto trucks for shipping to smelters or other recycled materials processing facilities.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Recycling and waste management is a low-digitization, small-firm-dominated sector with limited capital investment in automation. Public reports show minimal deployment of autonomous loading equipment in recycling operations compared to warehousing or manufacturing, with adoption remaining experimental rather than production-scale. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste and recycling is a physically-intensive, low-digitization sector with minimal AI/autonomy adoption in equipment operation compared to information-sector benchmarks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Current AI/automation offers limited augmentation for forklift or loader operation; these are already operator-controlled tasks where real-time decision-making and safety judgment dominate. Assisted steering or load-verification systems could provide modest gains, but AI does not meaningfully transform productivity on this primarily manual, site-specific task. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Some assistive technologies (camera systems, load-sensing, guidance aids) can help operators but do not fundamentally transform productivity on this specific loading task today. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While autonomous forklifts and loaders exist in controlled warehouse environments, operating these vehicles in recycling facilities—which have variable layouts, unpredictable material positioning, safety hazards, and outdoor/unstructured conditions—remains beyond reliable end-to-end automation today. Significant human oversight would still be required, preventing the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Autonomous forklift/loader technology exists but is far from generalizable to messy, variable recycling yard environments with irregular materials; most of this task still requires human operation today.ID |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Safety liability is a significant barrier: regulatory bodies (OSHA) and insurance carriers impose strict requirements on autonomous heavy equipment operation, especially near workers. Facilities also have union contracts and worker-protection mandates that constrain substitution, and the outdoor/hazardous nature of recycling sites creates legal exposure if automation fails. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing barrier akin to a professional credential, but safety regulations, liability for heavy equipment operation, and insurance requirements create meaningful friction for full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Autonomous forklifts and loaders carry high capital costs ($100k–$500k+ per unit), require facility retrofitting, and need ongoing maintenance and remote operation oversight. For a task performed by workers at typical recycling wages ($30k–$40k annually), the all-in amortized cost per task instance remains substantially higher than human labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Autonomous material handling equipment requires significant capital investment, sensors, and site modification, making it costlier than a human operator in most current recycling facility contexts. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Autonomous material handling systems are deployed in some organized industrial settings, but recycling operations present unique challenges (irregular bale shapes, dusty/hazardous environments, dynamic site layouts) that deployed products handle poorly. Existing solutions lack the robustness and safety certification required for routine production use in most recycling facilities. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Autonomous forklifts and lifts are deployed in some controlled warehouse settings, but recycling yards with irregular bales, debris, and outdoor terrain are not yet addressed by mature production systems. |
Clean materials, such as metals, according to recycling requirements.
23CI 10–35 · exposure 13 · augmentation 25 · importance 4.0/5 · click for rater detail
Clean materials, such as metals, according to recycling requirements.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Recycling and reclamation is a capital-constrained, fragmented sector with many small and medium-sized operators. Adoption of automated cleaning systems remains limited to larger facilities; most operations still rely on manual labor due to upfront costs and integration complexity. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste management and recycling is a low-digitization, physically intensive sector with minimal AI agent adoption for hands-on material processing tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI computer vision can assist workers in identifying contamination or material type errors, but current tools offer limited real-time guidance during active cleaning work. The manual, hands-on nature of the task and lack of proven augmentation tools in production limit meaningful productivity gains for human workers. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with sorting guidance or contamination detection via computer vision, but offers little direct help with the physical cleaning action itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Current AI and robotics can handle some material sorting and surface cleaning in controlled settings, but the task demands real-time identification of diverse material types, contamination detection, and adaptive cleaning procedures based on recycling specifications. End-to-end automation with 50% time savings at equal quality requires robust computer vision and dexterous manipulation not yet reliably deployed at scale in recycling facilities. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manual cleaning task requiring handling, sorting, and physical scrubbing/washing of materials, which current AI systems cannot perform without embodied robotics that don't exist at scale for this purpose. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Recycling facilities operate under environmental and safety regulations that create some oversight requirements, though no strict licensing mandate requires human sign-off on cleaning per se. Facility layout constraints, worker safety around machinery, and organizational inertia present moderate friction to automation adoption. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal barriers exist, but the physical nature of the task and need for specialized equipment create practical friction against any automation, AI or otherwise. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Robotic systems capable of material handling and cleaning remain capital-intensive and require ongoing maintenance, operator oversight, and facility customization. Loaded costs typically exceed the wages of semi-skilled recycling workers, especially when integrated into existing workflows and accounting for downtime. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-based substitute for physical cleaning, so any hypothetical automation would require expensive specialized robotics far costlier than manual labor or existing mechanical washing equipment. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some robotic cleaning systems exist in research and limited pilot deployments, but no mature production systems reliably perform this task across the diversity of materials and contamination profiles found in real recycling streams. Material-specific cleaning requirements and variable input quality create narrow scope and material error rates in deployed solutions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical cleaning of recyclable metals; this remains a manual or mechanical (non-AI) process in industrial settings. |
Collect and sort recyclable construction materials, such as concrete, drywall, plastics, or wood, into containers.
20CI 5–35 · exposure 13 · augmentation 38 · importance 4.3/5 · click for rater detail
Collect and sort recyclable construction materials, such as concrete, drywall, plastics, or wood, into containers.
20| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption is concentrated in large centralized recycling facilities rather than on-site collection; construction and waste sectors lag behind information and professional services in AI automation, with most sites still using manual labor. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Construction and waste management are low-digitization, physically intensive sectors with minimal AI/robotics adoption for onsite material handling. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-powered computer vision could assist workers in identifying and categorizing materials more quickly, and robotic assists could reduce physical strain during heavy sorting tasks, though human judgment and adaptability remain essential. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Some computer vision tools can help identify material types in stationary sorting facilities, but they offer little assistance to workers doing mobile collection and sorting on active construction sites. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Collection and sorting of construction materials requires physical handling, visual identification, and spatial reasoning in variable environmental conditions. Current AI robotic systems exist but are narrow in scope, require significant setup per site, and lack the dexterity and adaptability to reliably handle mixed, irregular construction debris at 50% time savings. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical materials handling task requiring mobility across variable construction sites, identification of mixed debris, and manual lifting/sorting; no off-the-shelf AI system performs this end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Physical presence on construction sites, heavy lifting liability concerns, safety certification requirements, and the need for human judgment in identifying hazardous materials create organizational and regulatory friction that protect human workers from full substitution. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement, but physical site variability, safety regulations around heavy materials, and lack of infrastructure for automation create practical friction beyond mere preference for humans. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized robotic sorting hardware, installation, integration, and site-specific customization remain expensive relative to low-wage recycling labor; meaningful cost parity or advantage is not yet achieved in real deployments at scale. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Robotic systems capable of mobile collection and sorting of heterogeneous construction debris would require expensive hardware, sensors, and maintenance far exceeding low-wage manual labor costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While robotic sorting systems exist in controlled environments (e.g., recycling facilities with uniform inputs), real construction-site collection involves unpredictable material composition, contamination, and spatial constraints where deployed solutions show high error rates and narrow applicability. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously collects and sorts construction debris on job sites; robotic sorting exists only in controlled facility settings for narrow waste streams, not in general construction contexts. |
Operate balers to compress recyclable materials into bundles or bales.
20CI 10–30 · exposure 13 · augmentation 25 · importance 4.1/5 · click for rater detail
Operate balers to compress recyclable materials into bundles or bales.
20| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Recycling facilities are typically small to medium operations with limited capital for automation investments. Adoption of AI-driven baler automation is minimal in practice; most facilities still rely on manual or semi-automatic operation by human workers. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste and recycling processing is a low-digitization, physical-labor sector with minimal AI/robotics adoption for this specific machine-operation task. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Current AI offers minimal augmentation for baler operators. Computer vision could potentially flag jamming or material quality issues, but such assistive features are not yet mainstream in recycling operations. The task is primarily mechanical and does not benefit substantially from partial AI support. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with sensor-based sorting or predictive maintenance scheduling around baler use, but offers little direct assistance to the physical act of operating the baler itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Baler operation involves sequential steps (feeding, compressing, tying) that are partially repetitive, but the task requires dynamic adjustment to material type, density, and jam clearance. Current AI and robots lack the dexterous manipulation and real-time feedback control needed for end-to-end autonomous operation without significant setup and human oversight. |
| Task automatability | claude-sonnet-5 | 1/5 | Physically operating and maintaining industrial baler equipment involves manual loading, monitoring, and troubleshooting that current AI systems cannot perform end-to-end without robotic embodiment.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Workplace safety regulations and OSHA standards require safeguards around heavy machinery; however, no specific legal mandate requires a licensed human to operate a baler, only compliance with equipment standards. Some organizational friction and safety oversight requirements exist but are not absolute blockers to automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement exists, but safety regulations, equipment liability, and physical workplace hazards create moderate friction against unsupervised automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Baler operation is a lower-wage task performed by recycling workers. The capital cost of a fully autonomous system, integration, and safety infrastructure would exceed the labor cost savings for most facilities. Human operators remain cheaper and more flexible. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-driven substitute for the physical task, so cost comparison favors the human operator by default since AI cannot perform the function. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed commercial systems autonomously operate balers in production environments. While some industrial automation exists for structured material handling, balers deal with variable, unstructured recyclable streams that demand human decision-making. Research prototypes exist but lack reliable real-world deployment at scale. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product operates balers autonomously; this remains a manual/mechanical operator task with existing automation limited to fixed-function machinery, not AI decision-making. |
Clean, inspect, or lubricate recyclable collection equipment or perform routine maintenance or minor repairs on recycling equipment, such as star gears, finger sorters, destoners, belts, and grinders.
19CI 15–24 · exposure 8 · augmentation 25 · importance 3.8/5 · click for rater detail
Clean, inspect, or lubricate recyclable collection equipment or perform routine maintenance or minor repairs on recycling equipment, such as star gears, finger sorters, destoners, belts, and grinders.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Recycling facilities are typically small to mid-sized operations with limited capital budgets and low digitization. There is minimal adoption of AI or advanced robotics in this sector; maintenance remains highly manual and labor-dependent with slow technology diffusion. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste management and recycling is a physically intensive, low-digitization sector with minimal AI/robotics adoption for hands-on maintenance tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by predictively monitoring equipment health and flagging maintenance needs, or providing remote diagnostic support to technicians, but current systems offer limited augmentation for the hands-on inspection and repair work that dominates this task. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with predictive maintenance scheduling or diagnostic alerts via sensors, but it offers little direct assistance to the physical cleaning and repair work itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While some aspects like visual inspection and logging could be partially automated with computer vision, the physical manipulation (lubricating, minor repairs, replacing parts like belts and gears) requires dexterous robotics that current systems cannot reliably perform in unstructured facility environments. The heterogeneous nature of equipment and failure modes limits end-to-end automation below the 50% threshold. |
| Task automatability | claude-sonnet-5 | 1/5 | This is hands-on physical inspection, cleaning, and mechanical repair of industrial equipment requiring manipulation, tactile feedback, and fine motor skills that current AI systems cannot perform without a capable robotic body. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While there are no hard licensing or legal barriers to automation, practical adoption is hindered by the need for custom engineering per equipment type, safety liability concerns with autonomous repair, and the technical difficulty of retrofitting existing facilities. Organizational friction is moderate but not absolute. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists specifically for this task, but safety protocols around lockout/tagout and heavy machinery create some procedural friction for any automation attempt. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Current robotic systems capable of maintenance tasks are extremely expensive to purchase, integrate, and maintain, far exceeding the cost of a recycling worker performing these routine tasks. The labor cost of a technician is substantially lower than the capital and operational cost of automation. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Robotic systems capable of this level of dexterous physical maintenance work are far more expensive to develop and deploy than paying a maintenance worker's wage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI or robotic systems currently perform routine maintenance and minor repairs on recycling equipment reliably in production environments. Visual inspection products exist in research, but physical intervention (lubrication, part replacement) remains absent from commercial deployment at scale. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs autonomous physical cleaning, lubrication, or minor repair of recycling machinery; this remains firmly in the domain of human maintenance workers. |
Operate automated refuse or manual recycling collection vehicles.
16CI 5–26 · exposure 13 · augmentation 25 · importance 3.6/5 · click for rater detail
Operate automated refuse or manual recycling collection vehicles.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Waste and recycling collection remains a traditional, often municipally or small-firm operated sector with limited digital infrastructure; adoption of autonomous collection vehicles is experimental and confined to a few pilots in major metros, not spreading at production scale. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste management is a low-digitization, physical-labor sector with minimal AI-driven displacement; the industry lags far behind information/professional services in automation adoption. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI assistance on this task is minimal; route optimization software exists but does not meaningfully augment the driver's core function of safely operating and positioning the vehicle and managing customer interactions on collection routes. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Some route optimization, scheduling software, and fleet telematics can improve efficiency, but there is little direct augmentation of the physical driving/collection task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While vehicle operation has autonomous capabilities in narrow environments (e.g., closed lots), recycling collection involves complex navigation of varied urban/rural terrain, interaction with obstacles, customer-specific bin placement, and safety-critical decisions that current autonomous systems cannot reliably handle end-to-end. Manual control remains essential for most routes. |
| Task automatability | claude-sonnet-5 | 1/5 | Physically driving and operating a collection vehicle through variable urban/suburban routes, handling obstacles and manual sorting requires embodied perception-action skills far beyond current AI capabilities.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Significant regulatory and liability barriers exist: vehicle operation is governed by commercial driving licenses and DOT regulations; autonomous heavy vehicles face insurance and legal uncertainty; municipalities and haulers face high error-cost asymmetry if a vehicle causes property damage or safety incidents. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing barrier akin to a doctor, but municipal contracts, safety regulations for heavy vehicles on public roads, and liability for accidents in populated areas create meaningful friction against autonomous operation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of developing, maintaining, and insuring autonomous collection vehicles with backup human oversight significantly exceeds the wage cost of a driver-operator, particularly when considering failure modes and the need for human intervention. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so any hypothetical automation (specialized robotics plus sensors) would be far more costly than paying a driver-operator today. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Autonomous refuse vehicles exist only in controlled pilots and limited geographies; no mature production system reliably handles general-purpose recycling collection routes with the variability of real-world streets, weather, and customer setups. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously operates refuse/recycling collection vehicles in production; autonomous driving in complex, low-speed, stop-and-go residential settings with manual loading remains research/pilot stage at best. |
Clean recycling yard by sweeping, raking, picking up broken glass and loose paper debris, or moving barrels and bins.
15CI 15–15 · exposure 0 · augmentation 13 · importance 4.1/5 · click for rater detail
Clean recycling yard by sweeping, raking, picking up broken glass and loose paper debris, or moving barrels and bins.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Recycling and reclamation is a lower-digitization sector with small firms and primarily manual operations; adoption of robotics for yard maintenance has been minimal and remains concentrated in research or pilot phases. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste management and recycling are low-digitization, physically intensive sectors with minimal AI/robotics adoption for manual yard-cleaning tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Current AI or robotics could provide limited augmentation (e.g., automated debris detection via computer vision to guide human cleanup), but practical assistance to a human performing sweeping and raking is minimal today. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers essentially no meaningful assistance to a worker sweeping, raking, or picking up debris by hand. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation of objects in unstructured outdoor environments, including sweeping, raking, and handling variable debris types. Current AI robots lack the dexterity, adaptability, and environmental perception to reliably perform these coordinated physical operations at scale. |
| Task automatability | claude-sonnet-5 | 1/5 | This is physical manual labor requiring mobility, dexterity, and perception in an unstructured outdoor environment; no off-the-shelf AI system performs this end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Recycling yard cleaning is low-barrier for automation legally, but the physical complexity and outdoor conditions create substantial practical friction. No licensing requirements prevent robotic substitution, but worker familiarity and organizational inertia are weak-to-moderate barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing or legal barriers restrict who cleans a yard, but physical/practical barriers (uneven terrain, hazardous debris, need for judgment) limit automation rather than regulation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Specialized robotics for outdoor yard cleaning would require significant capital investment, ongoing maintenance, and site customization—far exceeding the labor cost of recycling yard workers in most markets. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Cleaning robots capable of this variable, cluttered outdoor task do not exist commercially, so any hypothetical solution would be far more expensive than a human laborer. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial product reliably performs mixed yard-cleaning tasks (sweeping, picking up broken glass, moving barrels) in real recycling facilities. Specialized robotics exist for narrow subtasks but not end-to-end yard maintenance. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed robotic products reliably sweep, rake, and clear debris including broken glass in recycling yards at production scale; this remains far beyond current commercial robotics. |
Collect recyclable materials from curbside for delivery to designated facilities.
15CI 15–15 · exposure 0 · augmentation 13 · importance 3.6/5 · click for rater detail
Collect recyclable materials from curbside for delivery to designated facilities.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Recycling and reclamation is a labor-intensive, physical industry with minimal digital infrastructure; sector adoption of automation remains negligible and limited to research pilots. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Waste and recycling collection is a low-digitization, physically intensive sector with minimal AI/robotic adoption in real-world curbside operations. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | Current AI offers no meaningful assistance to the core task of physically collecting and transporting recyclable materials from curbside locations. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can assist with route optimization, scheduling, and truck logistics, but offers little direct assistance to the physical act of collecting materials. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical collection of materials from curbside locations and transportation to facilities—capabilities that current AI systems and even deployed robotics cannot perform reliably at scale in the unstructured outdoor environment of residential neighborhoods. |
| Task automatability | claude-sonnet-5 | 1/5 | This is physical labor requiring driving, lifting, and manual collection of materials from curbside locations, which current AI systems cannot perform end-to-end.:current AI has no embodied capability for this. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While there are no strict licensing barriers, the task involves public roadways, safety regulations, and local municipal contracts that create some organizational friction to automation adoption. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing barrier per se, but the physical nature of curbside collection, need for route judgment, and safety/liability concerns around municipal services create some friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Autonomous collection systems would require substantial capital investment in specialized vehicles and perception systems, along with high maintenance and supervision costs that vastly exceed the loaded wage of a collection worker. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI system substituting for this physical labor, so AI cost is not comparable or lower than human labor cost for this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial product performs curbside recycling collection end-to-end. Prototype robots exist in labs, but none operate reliably in production at the scale needed for municipal waste streams. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously collects curbside recyclables; this remains a manual or semi-automated (truck-assisted) human task. |
Extract chemicals from discarded appliances, such as air conditioners or refrigerators, using specialized machinery, such as refrigerant recovery equipment.
14CI 5–23 · exposure 13 · augmentation 25 · importance 4.2/5 · click for rater detail
Extract chemicals from discarded appliances, such as air conditioners or refrigerators, using specialized machinery, such as refrigerant recovery equipment.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Recycling facilities remain relatively low-tech and low-digitization environments with limited capital investment in advanced automation; adoption of AI or robotics in this sector is minimal and mostly confined to large industrial operations. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Recycling and waste management is a low-digitization, physical-labor sector with minimal AI/robotics adoption for hands-on hazardous material extraction. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with documentation, appliance classification, or safety alerts, but the core physical and chemical handling task offers limited scope for meaningful AI-driven productivity enhancement while keeping the worker in the loop. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with tracking inventory, compliance documentation, or scheduling, but offers little direct help with the physical extraction process itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While refrigerant recovery equipment is already specialized machinery, the task requires precise handling of hazardous chemicals, identification of different appliance types and refrigerant variants, and safe connection/disconnection procedures that demand dexterity, real-time decision-making, and safety judgment that current AI and robotics cannot reliably perform end-to-end without human oversight. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manipulation task requiring handling equipment, connecting recovery machines to appliances, and managing hazardous refrigerants; no AI system can perform this physical extraction process. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strict regulatory requirements govern refrigerant handling and recovery (EPA regulations in the US, similar rules globally), requiring certified technicians and documented compliance; liability for environmental or worker harm creates strong legal and organizational barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | EPA regulations (Section 608) require certified technicians to handle refrigerant recovery due to environmental hazards, creating a strong regulatory/licensing barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized refrigerant recovery equipment, robotic platforms capable of handling hazardous materials, and the integration/maintenance costs would exceed the loaded wage of recycling workers; automation here remains capital-intensive relative to low-wage manual labor. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute for this physical task, so AI cost is not comparable; human labor with specialized equipment remains the only viable option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed commercial AI system can autonomously extract refrigerants from appliances with the required precision and safety compliance; specialized robotics exist in limited industrial contexts but not as general-purpose, reliable production systems for this specific task at scale. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical refrigerant recovery; this remains a manual task using specialized recovery machinery operated by trained workers. |
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.