Embodied work
These systems touch vehicles, fields, production lines, infrastructure, laboratories, or material movement. Their authority is constrained by physics, safety envelopes, local operators, and fallback procedures that cannot be abstracted away.
Inclusion criteria
Includes cases with physical-work language in the industry, workflow, AI role, or operating description, including vehicles, equipment, fields, facilities, infrastructure, manufacturing, and material movement.
28 cases · bundle 1.0.0 · reviewed 2026-08-23
01
John Deere (Deere & Company)
Editorial leadThat a sprayer cannot tell a crop plant from a weed while moving through a field, so the only reliable way to control weeds is to broadcast herbicide across the entire field regardless of where weeds actually are.
Current · updated
1 independent, 1 primary; publication outcomes are verified and reported.
Canonical case →02
Rio Tinto Iron Ore
Editorial leadMainline driving can be automated while exceptions move to remote operations.
03
Amazon
Editorial leadHinge: multiple specialized robots can coordinate inventory presentation, package handling, and cart movement while employees shift toward quality, flow, maintenance, and exception work.
04
DHL Supply Chain
Editorial leadThat people must spend entire shifts lifting loose-loaded cases from hot or cold trailers.
05
Waymo
Editorial leadWithin a defined domain, automated driving can control the vehicle while remote assistance handles exceptions.
06
Johns Hopkins Children's Center / Johns Hopkins Medicine
Closing the diabetic eye-screening care gap requires persuading youth to attend a separate specialist appointment. Instead, the diagnostic exam itself is brought into the routine diabetes visit and read autonomously by AI in under a minute.
07
Meta (Facebook)
That every code change must run every test that transitively depends on modified files in order to catch regressions before trunk. Meta's predictive model estimates each test's likelihood of catching a regression for a specific change, so a much smaller learned subset can be run without loss of regression signal.
08
Swiss Plasma Center at EPFL and Google DeepMind
Operators set plasma objectives while a simulator-trained policy coordinates all coils on physical hardware.
09
Klarna
That a first-line customer-service errand requires a human agent to read, interpret and resolve it, so resolution capacity can only grow by adding agent headcount and outsourced capacity market by market.
Current · updated
2 independent, 1 primary; publication outcomes are verified and reported.
Canonical case →10
Twenty-eight South-East Queensland long-term care facilities
Hinge: a responsive social robot can provide repeatable short engagement sessions as an adjunct to care, without replacing clinical assessment or human relationships.
Watch status: verify the cited source and deployment condition before reusing this case.
Watch · updated
Watch status: verify the cited source and deployment condition before reusing this case.
1 peer reviewed; publication outcomes are verified.
Canonical case →11
TD Bank Group
That preparing a mortgage or HELOC file for a credit decision requires a human to read the document package end to end - classifying documents, extracting figures, calculating income, checking consents and writing the summary - because borrower documents are too variable for a rules engine to handle reliably.
Current · updated
1 independent, 1 primary; publication outcomes are verified and reported.
Canonical case →12
American Airlines
Live flight, runway, and gate data can recommend the nearest gate; operations teams retain disruption control.
13
Deutsche Bahn
AI can flag deviations from train images; technicians validate findings and trigger work before manual repair.
14
Ocado Group / Ocado Retail
Robotic arms can pick more grocery items on-grid while people handle exceptions and fulfillment operations.
15
Google DeepMind and EMBL-EBI
Researchers can use precomputed structure predictions, then decide what experimental validation is required.
16
Audi Neckarsulm
Quality can move from ultrasound sampling to AI screening of every weld.
17
Lawrence Berkeley National Laboratory and University of California, Berkeley
That scientists must manually choose every recipe, execute every synthesis step, interpret each XRD result, and plan the next experiment.
18
GitHub Copilot research experiment
Inline generation can compress a bounded task but not prove lifecycle productivity.
19
That AI should only recommend cooling set-point changes for operators to implement manually.
Current · updated
1 independent, 1 primary; publication outcomes are verified and reported.
Canonical case →20
NASA JPL
Communication latency requires local path authority.
21
Octopus Energy
Human agents must draft responses to all customer emails.
Current · updated
1 independent, 1 primary; publication outcomes are verified and reported.
Canonical case →22
City of Pittsburgh and Carnegie Mellon University
That urban intersections should follow fixed timing plans optimized offline for predictable rush periods.
Current · updated
1 independent, 1 peer reviewed; publication outcomes are verified.
Canonical case →23
Rwanda national blood service and Zipline
Remote hospitals can request blood without holding excess local stock or waiting for regional road transport.
24
Microsoft Azure
A cross-service Azure incident is recognized when a human on-call, seeing only a partial service view, pieces together cascading alerts and declares the incident.
25
National Grid Electricity Transmission
Central BVLOS drones/AI automate routine pylon imaging and corrosion grading; engineers decide maintenance.
26
Rio Tinto
A central system can drive mine trucks on fixed routes while people supervise and manage exceptions remotely.
Watch status: verify the cited source and deployment condition before reusing this case.
Watch · updated
Watch status: verify the cited source and deployment condition before reusing this case.
1 primary; publication outcomes are reported.
Canonical case →27
MIT randomized professional-writing study
The redesign is a shorter drafting loop with human revision and accountability, not autonomous publication.
Current · updated
1 independent, 1 peer reviewed; publication outcomes are verified.
Canonical case →28
New York City Fire Department
Fire companies can prioritize inspections by predicted risk instead of fixed cycles and local knowledge.
Current · updated
2 independent, 1 primary; publication outcomes are verified and reported.
Canonical case →