Crossing Guards and Flaggers

33-9091.00
Median wage $38,100/yr94,360 employed (US)Rank #898 of 923 scored · top 97% by substitution

Guide or control vehicular or pedestrian traffic at such places as streets, schools, railroad crossings, or construction sites.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution8
Exposure7
Augmentation22

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

12 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

0%

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%8

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

Technical feasibility todayw 20%5

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

Cost vs. human wagew 15%6

panel mean rating 1.2/5 → substitution pressure 6/100

Adoption barriersw 20%inverted — strong barriers lower the score16

panel mean rating 4.3/5 (barrier strength) → substitution pressure 16/100

Sector adoption velocityw 10%1

panel mean rating 1.0/5 → substitution pressure 1/100

Task breakdown (12 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.

Distribute traffic control signs and markers at designated points.

29

CI 5–54 · exposure 36 · augmentation 25 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Crossing guard and flagger positions are concentrated in public agencies, school districts, and small construction firms—sectors with low digitization, high regulatory compliance overhead, and limited capital budgets for automation. Adoption of robotic sign placement is minimal in these sectors.
Sector adoption velocityclaude-sonnet-51/5This occupation is physical, low-digitization, and part of a sector (traffic control/construction) with minimal AI or robotic adoption for manual placement tasks.
Augmentation potentialclaude-haiku-4-5-202510012/5AI or automation could assist by route-planning or sign inventory management, but the core task of physical placement at designated points offers limited augmentation value while a human remains in the loop; the task is inherently physical and location-specific.
Augmentation potentialclaude-sonnet-52/5AI could assist with route planning, optimal placement mapping, or scheduling, but offers little help with the core physical act of carrying and placing signs.
Task automatabilityclaude-haiku-4-5-202510015/5This task involves physically placing standardized signs and markers at predetermined locations. Autonomous systems (wheeled or aerial robots equipped with mounting hardware) could perform this end-to-end with >50% time savings by eliminating the need for human travel time, positioning checks, and physical placement labor.
Task automatabilityclaude-sonnet-51/5This is a physical task requiring transporting and placing heavy signs/cones at specific outdoor locations, which current AI systems (software-based) cannot perform without embodied robotics, which are not deployed for this purpose.
Adoption barriersclaude-haiku-4-5-202510014/5Traffic control sign placement often falls under municipal or DOT authority with specific procedural and liability requirements; human flaggers are frequently mandated by local traffic codes and OSHA regulations for safety compliance. Legal and regulatory barriers substantially protect this labor.
Adoption barriersclaude-sonnet-53/5While not licensed like a skilled trade, this task has practical/physical barriers (safety, terrain, regulatory placement standards) that require a present human, though no strict licensing regime applies specifically to sign placement.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current autonomous systems capable of sign distribution require significant capital investment, maintenance, and operational overhead. For sporadic or seasonal deployment (typical of traffic control), the all-in cost per task remains higher than hiring a minimum-wage crossing guard or flagger.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI/robotic substitute at any cost currently deployed for physically placing traffic signs, so AI cost is effectively infinite relative to a human doing the task.
Technical feasibility todayclaude-haiku-4-5-202510012/5While robotic systems exist for specialized placement tasks, no mature, widely-deployed products reliably perform sign and marker distribution at scale in real-world conditions. Prototypes exist but face challenges with varying terrain, weather, and exact placement requirements; production deployments remain rare.
Technical feasibility todayclaude-sonnet-51/5No deployed products autonomously distribute physical traffic control signs and markers; this remains manual labor done by human workers or crews.

Record license numbers of vehicles disregarding traffic signals, and report infractions to appropriate authorities.

17

CI 9–25 · exposure 17 · augmentation 38 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Crossing guard and traffic enforcement roles are typically municipal/government functions in small towns and suburban areas with low digitization and high human-preference-for-presence requirements; adoption of automated enforcement is limited by regulation and budget constraints.
Sector adoption velocityclaude-sonnet-52/5Traffic control and crossing guard roles are in a low-digitization, physical-presence sector where AI camera adoption is slow and municipality-dependent.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by flagging suspicious vehicle activity or highlighting license plates in video for human review, but current systems are error-prone enough that meaningful augmentation is limited; a crossing guard still performs the core task of watching, judgment, and official reporting.
Augmentation potentialclaude-sonnet-53/5Where ALPR cameras are installed, they can assist guards by automatically capturing plate data, reducing manual recording effort, though this requires existing infrastructure not typically tied to individual guards.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time visual identification of vehicles, reading license plates at variable angles and distances, and legal judgment about traffic signal compliance—capabilities that current AI systems cannot reliably perform in uncontrolled outdoor environments at the speed and accuracy needed for enforcement.
Task automatabilityclaude-sonnet-52/5Recording plate numbers and reporting infractions requires real-time visual monitoring and physical presence at a location, which current AI systems cannot autonomously perform without dedicated camera infrastructure already in place.:
Adoption barriersclaude-haiku-4-5-202510015/5Traffic enforcement and the legal authority to issue or report citations are tightly regulated; only authorized municipal or law-enforcement personnel can legally record violations and report them to authorities, creating a hard legal and liability barrier to AI automation.
Adoption barriersclaude-sonnet-54/5Traffic enforcement often requires certified equipment, chain-of-custody for citations, and legal authority to issue tickets, creating regulatory and liability barriers to full automation of this specific reporting task.
Cost vs. human wageclaude-haiku-4-5-202510012/5Fixed camera systems with LPR and human review operators are expensive to deploy and maintain across multiple intersections, and the human oversight required to filter false positives means total cost remains comparable to or higher than paying crossing guards directly.
Cost vs. human wageclaude-sonnet-52/5Deploying ALPR camera systems with associated infrastructure, maintenance, and legal review is costly compared to a guard simply noting plates, though at scale cameras can be cheaper per-incident.
Technical feasibility todayclaude-haiku-4-5-202510012/5While ANSI cameras and license plate recognition (LPR) systems exist in limited deployments, they have material error rates in variable lighting, plate angles, and occlusion; moreover, they require fixed infrastructure and human verification, making autonomous end-to-end recording and reporting unreliable in the field conditions crossing guards face.
Technical feasibility todayclaude-sonnet-52/5Automated license plate recognition and traffic violation cameras exist and are deployed in some jurisdictions, but they are separate fixed infrastructure systems, not a substitute for the guard's in-person observational and reporting role.

Discuss traffic routing plans and control-point locations with superiors.

14

CI 5–23 · exposure 8 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Traffic control and crossing guard operations are conducted by lower-tech, highly localized organizations with minimal AI infrastructure and strong human-safety-based decision-making norms.
Sector adoption velocityclaude-sonnet-51/5Crossing guard and flagger work is a low-digitization, physical, small-scale occupational sector with minimal AI adoption in daily operational planning.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by summarizing routing data or drafting proposals, but the core task—discussing and resolving plans with superiors—remains fundamentally collaborative and human-centered.
Augmentation potentialclaude-sonnet-53/5AI mapping and traffic-simulation tools can help prepare data, visualize routes, and draft proposals, usefully augmenting the human-led discussion with a superior.
Task automatabilityclaude-haiku-4-5-202510011/5This task involves interpersonal discussion and collaborative planning with supervisors, requiring contextual judgment about traffic scenarios and organizational constraints that current AI cannot perform autonomously.
Task automatabilityclaude-sonnet-52/5This involves interactive discussion, situational judgment about local traffic conditions, and negotiation with a supervisor that current AI cannot fully replicate end-to-end.,though AI could help draft or summarize plans.,
Adoption barriersclaude-haiku-4-5-202510014/5Human judgment, accountability, and supervisory sign-off are essential to traffic safety operations; regulatory and liability frameworks expect human personnel to be responsible for control-point decisions.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement blocks AI involvement in planning discussions, but organizational hierarchy and safety-critical decision-making create moderate friction to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of an AI system to facilitate or replace this conversational task would exceed the loaded wage of a crossing guard or flagger having a brief supervisory discussion.
Cost vs. human wageclaude-sonnet-52/5Since AI cannot perform the core interactive discussion role, any cost comparison favors the human doing the actual conversation, though AI-assisted planning tools could lower prep costs somewhat.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can reliably conduct autonomous discussions with human supervisors about traffic routing; this requires human judgment and relationship management that is not productized.
Technical feasibility todayclaude-sonnet-51/5No deployed product exists that autonomously discusses traffic routing plans and control-point placement with a human superior in this occupational context.

Guide or control vehicular or pedestrian traffic at such places as street and railroad crossings and construction sites.

11

CI 0–23 · exposure 13 · augmentation 13 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI traffic automation remains minimal; schools and construction sites still rely almost entirely on human crossing guards and flaggers, with limited pilot programs and slow institutional change in this low-digitization, safety-critical domain.
Sector adoption velocityclaude-sonnet-51/5This occupation is in a low-digitization, physical-labor sector with essentially no AI agent adoption; traffic control automation (like sensors) has been static for decades, not part of a fast AI adoption trend.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with vehicle detection or alert systems, but the core task—making real-time safety decisions and issuing commands to live traffic—requires human judgment and accountability, limiting augmentation potential beyond modest sensor support.
Augmentation potentialclaude-sonnet-51/5AI offers negligible assistance to the core task of standing at a location and physically signaling traffic; there is little tech augmentation applicable to the moment-to-moment execution of this role.
Task automatabilityclaude-haiku-4-5-202510012/5Automating real-time traffic control in mixed environments requires reliable perception of dynamic actors (vehicles, pedestrians, cyclists) and safety-critical decision-making. While computer vision can detect traffic patterns, current AI cannot safely assume control of live traffic guidance without extensive, human-supervised infrastructure integration, leaving only narrow automation of simple signaling.
Task automatabilityclaude-sonnet-51/5This is a physical, real-time task requiring presence at a specific location to direct human drivers and pedestrians, which current AI (software/LLM-based) cannot perform end-to-end; it requires physical embodiment and situational judgment in unpredictable environments.
Adoption barriersclaude-haiku-4-5-202510015/5Traffic control is heavily regulated and often legally mandated to be performed by trained, licensed personnel (crossing guards, flaggers, traffic officers); liability for accidents or improper guidance creates strong legal and organizational barriers to full AI substitution.
Adoption barriersclaude-sonnet-54/5Safety regulations, liability for traffic accidents, and requirements for certified/trained personnel at construction sites and railroad crossings create strong barriers against automation despite no formal licensing exam typically required.
Cost vs. human wageclaude-haiku-4-5-202510012/5Crossing guard wages are relatively modest, and the infrastructure, sensors, and maintenance required to automate traffic control safely—plus ongoing human oversight—likely exceed the cost of a single human guard in most jurisdictions.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical task, so any comparison is moot—robotic solutions would be far more costly than a human worker with a sign and vest.
Technical feasibility todayclaude-haiku-4-5-202510012/5Deployed systems can detect vehicles and pedestrians in controlled contexts, but no mainstream product reliably manages live traffic control at scale without human oversight. Pilot autonomous traffic systems exist in limited lab settings, but production traffic guidance still depends on trained humans or fixed signals.
Technical feasibility todayclaude-sonnet-51/5No deployed product replaces a human crossing guard or flagger holding a sign and physically directing traffic; automated signage and traffic lights exist but are not equivalent AI agents performing this task.

Monitor traffic flow to locate safe gaps through which pedestrians can cross streets.

11

CI 0–23 · exposure 13 · augmentation 25 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Crossing guard work is concentrated in school districts and municipal/government roles—traditionally conservative, low-digitization sectors with high human-contact requirements and low automation appetite. No measurable displacement or production AI adoption in this function is evident.
Sector adoption velocityclaude-sonnet-51/5This occupation sits in a low-digitization, physical safety sector with essentially no AI/robotic adoption in production for this specific task.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist a human crossing guard by highlighting vehicle presence or suggesting safer timing, but the task is fundamentally observational and real-time, leaving limited space for AI to substantially amplify human productivity without the human already being on-site performing core monitoring.
Augmentation potentialclaude-sonnet-52/5Smart traffic sensors or AI-based traffic signal systems could provide supplementary data, but they offer minimal practical assistance to a human guard's moment-to-moment judgment and physical positioning.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI systems struggle with real-time multi-modal traffic monitoring and pedestrian safety assessment. While computer vision can detect vehicles and gaps, the task requires dynamic judgment under variable conditions (weather, speed, sight lines) and legal/safety accountability that AI cannot reliably assume end-to-end without extensive human oversight.
Task automatabilityclaude-sonnet-51/5This requires real-time physical presence, split-second visual judgment, and physical positioning to protect pedestrians in dynamic traffic conditions; no off-the-shelf AI system performs this embodied real-world function end-to-end today.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: crossing guards often work near schools or vulnerable populations under local ordinances and DOT regulations that may mandate human presence; liability for failed crossing decisions is high; and public/parental preference for human vigilance near children creates regulatory and social friction against full automation.
Adoption barriersclaude-sonnet-55/5Physical presence, legal/liability requirements for child and pedestrian safety, and municipal/school regulations mandating human crossing guards create hard barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5A crossing guard's loaded wage (~$25–35K annually) is low-cost manual labor. Deploying computer vision infrastructure, real-time processing, communication to pedestrians, and 24/7 operation would exceed the savings from replacing a single guard, especially given liability and oversight needs.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical safety task, so any comparison would require robotic hardware plus sensors far exceeding the cost of a human guard's wage.
Technical feasibility todayclaude-haiku-4-5-202510012/5Deployed AI can identify vehicles and measure gaps in controlled settings, but no production system reliably monitors live traffic to make authoritative crossing safety calls for pedestrians. Camera-based gap detection exists in research; deployment in actual street crossing scenarios with liability remains limited.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously performs live pedestrian crossing guidance in real traffic; this remains a physical safety role requiring human presence and liability accountability.

Learn the location and purpose of street traffic signs within assigned patrol areas.

5

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Crossing guard work is a physical, on-site public safety role with low digitization and minimal AI adoption. This specific learning task occurs in low-tech, municipal or school settings where automation uptake is typically lagging.
Sector adoption velocityclaude-sonnet-51/5This occupation is physical, low-digitization, and shows minimal AI adoption in practice; it is a laggard-sector task.
Augmentation potentialclaude-haiku-4-5-202510012/5An AI system could marginally assist by providing a digital map or database of traffic signs in the patrol area, but the core task of learning and internalizing this information remains primarily human-driven. The augmentation potential is limited and low-impact.
Augmentation potentialclaude-sonnet-52/5AI could provide reference maps or sign databases to aid onboarding, but this offers only marginal assistance to a task fundamentally learned through direct physical familiarization.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires on-site learning of specific local street infrastructure, which is primarily a human memory and observation task. While an AI could theoretically catalog traffic sign locations from imagery or databases, the core task—actively learning and internalizing the location and purpose of signs within a specific patrol area through assignment—is fundamentally a human cognitive process that offers no meaningful efficiency gain through automation.
Task automatabilityclaude-sonnet-51/5This requires physical presence, spatial memorization, and real-world observation at a specific site; no AI system can physically learn a patrol area's signage layout in a way that substitutes for the human role.'
Adoption barriersclaude-haiku-4-5-202510014/5Crossing guards are legally responsible for their patrol areas and must be familiar with local conditions to perform their safety-critical duties. Legal and liability requirements mean a human must personally learn and validate the conditions in their assigned area, creating a hard barrier to automation.
Adoption barriersclaude-sonnet-54/5Crossing guard duties involve public safety and often require in-person presence, local authorization, and physical positioning, creating strong practical barriers to remote or automated substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5An AI system capable of cataloging and organizing traffic sign data would require specialized deployment, data collection, and integration—all expensive relative to the minimal human labor cost of a crossing guard reviewing their assigned area and learning sign locations during initial onboarding.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so any AI cost is irrelevant relative to the human doing it as part of routine duty.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs this task end-to-end. Although AI can analyze street imagery or read traffic sign classifications, the actual learning of a specific patrol area's signs by an individual crossing guard is a knowledge internalization task for which no production system is designed.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs this on-site learning task for crossing guards; it is inherently tied to physical embodiment and local knowledge acquisition.

Direct or escort pedestrians across streets, stopping traffic, as necessary.

3

CI 0–5 · exposure 0 · augmentation 0 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI in this task is negligible; infrastructure and regulatory frameworks still require human crossing guards, particularly at schools and high-traffic intersections.
Sector adoption velocityclaude-sonnet-51/5This occupation is in a low-digitization, physical-labor sector (municipal/school safety services) showing essentially no AI adoption or displacement trends.
Augmentation potentialclaude-haiku-4-5-202510011/5AI offers minimal assistance to crossing guards performing this task; the core work—physically stopping traffic and ensuring pedestrian safety—depends entirely on human presence and judgment.
Augmentation potentialclaude-sonnet-51/5AI offers negligible assistance to the core physical act of stopping traffic and escorting pedestrians; smart traffic signals or sensors are adjacent infrastructure, not augmentation of the guard's task.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires physical presence on streets to stop traffic and guide pedestrians, involving real-time interaction with drivers and pedestrians in unpredictable street environments. No current AI system can physically intercept vehicles or embodied presence without human control.
Task automatabilityclaude-sonnet-51/5This requires real-time physical presence, judgment about traffic conditions, and physical intervention (stepping into streets, holding signs, physically guiding people); no off-the-shelf AI can perform this physical act end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5This task is legally protected: crossing guards and flaggers are required by regulation in many jurisdictions, and liability for accidents falls on authorized personnel. Human presence and legal authorization are hard requirements.
Adoption barriersclaude-sonnet-54/5Liability for child/pedestrian safety, physical presence requirements, and municipal/school district policies mandating human staffing create strong practical and quasi-regulatory barriers to automation, even though it isn't a formally licensed profession.
Cost vs. human wageclaude-haiku-4-5-202510011/5Autonomous systems capable of physical traffic control would require expensive hardware, infrastructure modification, and continuous operation far exceeding the loaded wage of a crossing guard.
Cost vs. human wageclaude-sonnet-51/5Replacing a human with an AI-capable robot or system for physical street-crossing supervision would require expensive robotics and sensors, likely costing far more than a minimum-wage crossing guard, with no viable products at any cost today.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can autonomously direct pedestrians across streets or stop traffic. The task requires embodied presence, real-time judgment of traffic patterns, and legal authority that current AI systems cannot provide.
Technical feasibility todayclaude-sonnet-51/5No deployed product physically escorts pedestrians or stops traffic in place of a human; automated traffic signals exist but do not replace the escort/guard function, which remains fully human-performed.

Communicate traffic and crossing rules and other information to students and adults.

3

CI 0–5 · exposure 0 · augmentation 13 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Crossing guard and flagger roles remain almost entirely human-performed; there is minimal evidence of AI or automation adoption in this occupation, which is concentrated in low-tech, public-sector, and school environments with limited digitization.
Sector adoption velocityclaude-sonnet-51/5This occupation is physical, low-digitization, and government/municipal-run, a laggard sector with minimal AI adoption in real-time field operations.
Augmentation potentialclaude-haiku-4-5-202510011/5AI offers negligible assistance to a crossing guard's core task of communicating safety rules to pedestrians. While real-time traffic detection might inform a human, the interpersonal and authoritative communication role is not meaningfully augmented by current systems.
Augmentation potentialclaude-sonnet-52/5AI could help prepare training materials or communicate rules via apps/signage to supplement the guard's messaging, but this offers only marginal assistance to the core in-person task.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time, context-aware verbal or gestural communication with pedestrians in dynamic traffic situations. Current AI systems cannot reliably perceive complex street conditions, judge when to intervene, and deliver appropriate safety instructions to diverse populations on-site, making end-to-end automation infeasible.
Task automatabilityclaude-sonnet-51/5This requires real-time physical presence, situational awareness of moving vehicles and pedestrians, and immediate verbal/gestural communication on-site, which current AI cannot perform end-to-end. Only trivial informational sub-components (e.g., producing a handout) are automatable.
Adoption barriersclaude-haiku-4-5-202510015/5Many jurisdictions legally require a human crossing guard or certified flagger to manage pedestrian safety at crossings. Liability and duty-of-care requirements mean that a responsible adult human must be present and accountable, creating a hard legal and institutional barrier to automation.
Adoption barriersclaude-sonnet-54/5Physical safety liability, legal requirements for trained personnel controlling traffic near schools, and the need for real-time human judgment and authority create strong barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure cost of deploying autonomous systems (sensors, compute, communication, maintenance) to replace a human crossing guard would far exceed the modest wage of the role, especially given the safety-critical nature requiring redundancy and oversight.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical, real-time safety task, so any AI solution (e.g., robotic signage) would require costly hardware and infrastructure far exceeding a human guard's wage for equivalent reliability.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs the role of a crossing guard or flagger in production. While AI vision systems can detect traffic patterns, there is no integrated system reliably communicating safety rules to pedestrians in real-world conditions at scale.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product stands at a physical crossing directing real-time traffic communication to students and adults; this remains outside current product capability.

Direct traffic movement or warn of hazards, using signs, flags, lanterns, and hand signals.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This task is concentrated in lower-digitization sectors (schools, construction, municipalities) with strong institutional attachment to human presence for child and public safety. Adoption of AI traffic direction is virtually absent; municipalities have no incentive to replace safety-critical human judgment.
Sector adoption velocityclaude-sonnet-51/5This occupation is in a highly physical, low-digitization sector with essentially no AI or robotic adoption for the core task.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist by alerting flaggers to detected hazards via computer vision (e.g., vehicles approaching unsafely), but the core task of physical signaling and real-time judgment requires the human to remain fully in control. Assistance value is limited and not yet deployed at scale.
Augmentation potentialclaude-sonnet-52/5AI could assist with scheduling, traffic pattern analysis, or hazard alerts via sensors, but offers minimal direct assistance to the moment-to-moment physical act of directing traffic.
Task automatabilityclaude-haiku-4-5-202510011/5Directing traffic and warning of hazards requires real-time perception of complex road conditions, dynamic decision-making about vehicle and pedestrian movement, and physical presence with flags/signs at the location. Current AI systems cannot reliably perform these safety-critical tasks end-to-end in unpredictable physical environments.
Task automatabilityclaude-sonnet-51/5This requires real-time physical presence, situational awareness of vehicles and pedestrians, and split-second decision-making in dynamic outdoor environments that current AI cannot perform end-to-end without robotic embodiment.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and regulatory barriers exist: traffic safety laws typically require licensed or certified human personnel in active traffic direction roles, and liability for traffic accidents and pedestrian safety fall on the responsible organization. Public safety and liability create hard adoption barriers.
Adoption barriersclaude-sonnet-54/5Liability for pedestrian and worker safety, legal requirements for trained personnel at school crossings and work zones, and physical presence requirements create strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The infrastructure cost (autonomous signaling hardware, robust sensors, 24/7 operation, liability coverage) would far exceed the wage of a crossing guard or flagger, especially given the safety-critical nature and low task density in many locations.
Cost vs. human wageclaude-sonnet-51/5No viable AI substitute exists for the physical act, so any comparison favors the human worker who is currently far cheaper than hypothetical robotic alternatives.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs traffic direction or hazard warning in the field as a substitute for human crossing guards or flaggers. While computer vision can detect hazards, integrating it with physical signaling devices and real-time traffic management remains research-stage.
Technical feasibility todayclaude-sonnet-51/5There are no deployed AI products that physically stand in roads directing traffic with signs and hand signals; this remains outside current product capability entirely.

Report unsafe behavior of children to school officials.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Schools operate in a heavily regulated, human-centered safety context with legal liability concerns and trust norms that strongly favor human staff for child observation and reporting. Adoption of AI for this task is negligible, and sector digitization does not meaningfully advance AI replacement here.
Sector adoption velocityclaude-sonnet-51/5Crossing guard work is a low-digitization, physical, low-wage occupation with essentially no AI adoption trend in this specific reporting function.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist a crossing guard by flagging anomalies (e.g., unusual congestion or potential hazards in video footage), but the core task—interpreting child behavior and judging whether it is unsafe—remains fundamentally a human responsibility, limiting augmentation value.
Augmentation potentialclaude-sonnet-52/5Basic video monitoring or alert systems could theoretically flag anomalies for human review, but no meaningful current tool assists crossing guards in observing and reporting child behavior.
Task automatabilityclaude-haiku-4-5-202510011/5Reporting unsafe behavior requires subjective judgment about intent, context, and severity—judgments that depend on social and developmental understanding children possess. Current AI cannot reliably interpret child behavior nuance or make the nuanced safety assessments a human observer can, let alone replace the human judgment required for accurate reporting.
Task automatabilityclaude-sonnet-51/5This requires physical presence at a crossing, real-time observation of children's behavior, and human judgment about safety followed by interpersonal communication with school officials; no AI system performs this end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Strong barriers exist: schools have duty-of-care and liability obligations that generally require a responsible human adult to observe and report; parents and regulators expect human judgment and accountability in child safety reporting; legal and organizational frameworks center on human personnel making these judgments.
Adoption barriersclaude-sonnet-54/5Child safety monitoring involves strong liability concerns, privacy/surveillance regulations around minors, and expectations of human judgment and accountability, creating substantial barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of vision systems, integration, human oversight, and error liability would far exceed the modest wage of a crossing guard performing this occasional task. Errors in reporting behavior could trigger costly false reports or miss genuine safety issues.
Cost vs. human wageclaude-sonnet-51/5Any AI solution would require costly camera infrastructure, computer vision monitoring, and human oversight, likely exceeding the low wage cost of a human crossing guard performing this incidental task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably identifies and reports unsafe child behavior to school officials in production. This task requires real-time observation, contextual reasoning, and integration with school incident-reporting systems that no current AI system performs end-to-end.
Technical feasibility todayclaude-sonnet-51/5No deployed product monitors street crossings for unsafe child behavior and autonomously reports to school officials; this remains outside current commercial AI capability.

Inform drivers of detour routes through construction sites.

3

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This is a physical, site-specific task in construction and transportation sectors with low digitization and high human-contact requirements; adoption remains near zero despite decades of automation opportunity.
Sector adoption velocityclaude-sonnet-51/5Construction and traffic control sectors have very low AI/digitization adoption for physical, real-time human safety tasks like this.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist with traffic pattern monitoring or route planning offline, but provides minimal productivity gain for the real-time communication and judgment demands of the active task itself.
Augmentation potentialclaude-sonnet-52/5AI-powered signage, smart traffic systems, or digital detour notifications could supplement information delivery, but they don't meaningfully augment the guard's core real-time physical signaling and communication role.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time interaction with drivers in variable physical environments, reading human behavior and traffic conditions, and adapting communication dynamically. Current AI cannot safely or reliably perform this end-to-end in the field.
Task automatabilityclaude-sonnet-51/5This requires physical presence to visually assess real-time traffic conditions, hold signage, and directly communicate with drivers at a physical location, which current AI systems cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Public safety roles directing traffic have strong legal and regulatory requirements for human presence and accountability; liability for accidents or injuries creates hard barriers to full automation without a licensed human responsible for safety.
Adoption barriersclaude-sonnet-54/5Traffic control often requires certified/trained personnel, physical presence for safety enforcement, and liability concerns around vehicle-pedestrian interactions create strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Even if feasible, the cost of deploying autonomous systems with safety infrastructure, liability coverage, and oversight would far exceed the loaded wage of a single crossing guard or flagger.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this physical, situational task, so no meaningful cost comparison favors AI; a human worker remains the only option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product today can independently manage live traffic communication at construction sites, interpret driver intent, and issue appropriate instructions to moving vehicles in real conditions.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product physically stands at construction sites directing traffic and verbally informing drivers of detours; this remains purely a human physical task.

Stop speeding vehicles to warn drivers of traffic laws.

0

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This task remains almost entirely performed by humans; adoption of automation is negligible, with only limited experiments in camera-based speed detection in some jurisdictions, not vehicle stopping.
Sector adoption velocityclaude-sonnet-51/5Physical, low-digitization occupations like crossing guards and flaggers show minimal AI adoption in real-world deployment for this specific physical task.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by analyzing traffic patterns or providing alerts about speeding vehicles, but cannot augment the core task of stopping vehicles and warning drivers in real time.
Augmentation potentialclaude-sonnet-52/5AI could assist with speed detection alerts or traffic pattern analysis feeding into guard positioning, but it offers little direct assistance to the moment-to-moment physical act of stopping vehicles.
Task automatabilityclaude-haiku-4-5-202510011/5Stopping vehicles and warning drivers requires physical presence at the roadside, interpersonal communication with drivers, and real-time judgment in dynamic traffic environments—capabilities far beyond current AI systems.
Task automatabilityclaude-sonnet-51/5This requires physical presence, real-time perception of oncoming vehicles, and authoritative human intervention on a public roadway; no current AI system can physically stop or confront a speeding vehicle.
Adoption barriersclaude-haiku-4-5-202510015/5Traffic enforcement and vehicle stopping require legal authority held only by licensed law enforcement or designated traffic control personnel; liability, safety, and regulatory frameworks place hard legal barriers on non-human enforcement.
Adoption barriersclaude-sonnet-55/5Physical safety enforcement on roads typically requires an authorized human (often with legal standing) to flag down and warn drivers; liability, safety, and legal authority create hard barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Any hypothetical automated enforcement system (cameras, sensors, etc.) would require massive infrastructure investment, integration, and maintenance costs far exceeding the wage of a crossing guard.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical function, so AI cost is not comparable—effectively infinitely more expensive since no product exists to do the job.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can physically intercept vehicles or conduct in-person traffic enforcement; this task is legally and operationally the domain of human traffic control workers.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs physical traffic stopping or in-person driver warnings; this remains entirely a human physical-presence task.

Related occupations — Protective Service

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.