OperatorsSite teamsControl roomsEnterprise systems
The teams and systems that commission work and receive the result.
WorkMachines / Physical AI operations
The operating layer for autonomous machines.
WorkMachines connects human operators, autonomous machines and real site systems in one accountable operation. Patrol is the first commercial product. Crop Monitoring is the next R&D programme on the same platform.
01Human authority
02Site-edge execution
03Evidence by default
WM
PatrolOperator console / illustrative
Active siteNorthbank Logistics
P-01 connected 74%
LiveP-01 / Front RGB
0.4 m/s 082° 74% 38 ms
Mission events3 requiring review
SolutionsOne operating layer. Distinct applications for real site work.
01 / In developmentPatrol02 / R&D programmeCrop MonitoringFuture applicationsInspection · Inventory · Infrastructure
Solution 01 / Patrol
Autonomous patrol, from route to verified handoff.
The robot covers the route. WorkMachines supervises the mission, checks what it finds and returns a complete incident record to the people already responsible.
Explore Patrol51.5082 N / 0.3411 W
Robot P-01Executing route
East loading gateObject detected
22:00Mission started
22:12Checkpoint 07
22:18Object detected
22:18Evidence package
Solution 02 / R&D programme
Crop Monitoring for protected growing environments.
A mobile machine follows repeatable routes, records crop and site conditions, and returns location-tagged observations for a grower to review.
Explore Crop MonitoringZONE PLAN / GH-A
Mobile unit M-01Recording row A-03
Zone A-04 / 14:32Variation recorded
The WorkMachines system
One platform, configured for different kinds of work.
Patrol and Crop Monitoring share mission control, site-edge execution and machine connectors. Their routes, sensing plans and review workflows remain specific to the job.
System architecture
The full system used by every WorkMachines application.
PatrolCrop monitoringInspectionFurther applications
Packaged workflows built on the same operating layer.
WorkMachines operating system
Mission orchestrationFleet stateIdentity & accessEvidence catalogue
Missions, fleet state and the complete operational record.
Perception servicesPlanning & agentsPolicy engineEvent verification
Perception becomes a governed decision before any machine action.
Mission runtimeLocal policyHealth & diagnosticsEvidence buffer
The active mission stays close to the machine and keeps running locally.
Capability contractCommand translationSDK / ROS bridgeState normalisation
Robot-specific commands and states sit behind a common contract.
QuadrupedsAMRs & dronesSensorsRoutes & infrastructure
Machines, sensors, routes, assets and infrastructure.
Commands move down Telemetry and evidence returnSite execution remains local when connectivity drops
Choose an application to see the parts it uses.Open the architecture
Advisory
Make the deployment decision before buying the machine.
We combine adoption and deployment planning into one path. Start with the unresolved decision and stop when the evidence is enough.
See Advisory01Evaluate
Find the site and task worth testing.
02Design
Specify the machine, controls and economics.
03Prove
Run the job under real site conditions.
04Operationalise
Set ownership, response and rollout gates.