M / 01
Hours-on-floor
Continuous operating time under realistic shift conditions, without engineering supervision.
Pilots
Pilots are not demos. They are deployments.
We accept a small number of pilots each quarter. We optimise for environments where the gap between prototype performance and production reality is largest, and where success compounds — into a unit that survives on site, and into the next deployment after that.
01 — Who should apply
We work with operators, not audiences.
The pilot is most productive when the applicant is the team that already owns the operating problem — not an exploratory project, not a marketing-led initiative.
/ 01
Industrial operators
Factories, sites, warehouses, hubs, care facilities, retail floors, hospitality venues with a defined bottleneck.
/ 02
Robotics firms
Teams shipping hardware that need a deployment-grade cognitive layer to reach production maturity.
/ 03
Deployment partners
Systems integrators and facility partners moving multiple platforms into the same operating environment.
02 — What pilots test
A measurable thing, not a feeling.
Each pilot defines a small number of operational metrics up front. We do not run open-ended discovery. We run timed, instrumented deployments against a specific bottleneck.
M / 02
Exception rate & recovery
Frequency of operational exceptions and the share resolved by the system without human intervention.
M / 03
Decisions audited
Non-trivial autonomous decisions with full reasoning trail, reviewable by the operator's supervisor.
M / 04
Operator time recovered
Hours of human attention freed from supervision and recovered for higher-value work.
03 — What environments matter
Where pilots make the most difference.
We give priority to environments that already have robotic platforms in place and a documented gap between current performance and operating expectation.
/ Env 01
High-mix manufacturing
Lines where SKU mix, jig wear, or upstream variability is breaking pre-programmed automation.
/ Env 02
Active construction sites
Sites where the gap between BIM and reality is large and where exception handling dominates time-on-task.
/ Env 03
Multi-fleet warehouses
Operations running multiple AMR/AGV platforms that need a shared reasoning layer above the WMS.
/ Env 04
Clinical-grade care
Care environments where reliability, auditability, and escalation discipline are non-negotiable.
/ Env 05
High-traffic retail
Stores where customer flow, language coverage, and live stock state collide.
/ Env 06
Mid-scale hospitality
Venues running multi-system coordination across reservation, ticketing, and operations.
04 — What we evaluate
A short, deliberate selection.
/ 01
Operational fit
Is the bottleneck genuinely cognitive, or is it a hardware or process problem dressed up as autonomy?
/ 02
Engineering readiness
Is there a counterpart team on the operator side who can co-own the integration?
/ 03
Time horizon
Is the operator's timeline measured in quarters, not press cycles?
05 — Apply
Open a pilot conversation.
One submission, reviewed by the deployment team. We respond if there is operational fit.