Product
One execution record for the entire physical job.
Kinemica connects the requested outcome to the workers, decisions, evidence and final result. Planning is useful, but permission and proof determine whether work can proceed and close.
Receive → verify
The job lifecycle
The job lifecycle
- 01
Job intake
Receive the requested outcome, source material, operating context, restrictions and expected evidence.
- 02
Planning
Produce a structured proposal with tasks, dependencies, risks, source references and uncertainties.
- 03
Allocation
Propose work across AI agents, human workers and robots using one worker and capability model.
- 04
Permission checks
Evaluate identity, status, qualification, capability, location, risk, dependencies and current state.
- 05
Execution
Start or dispatch only permitted work. Machine commands remain bounded and adapter-specific.
- 06
Evidence
Preserve original photographs, measurements, confirmations, approvals, telemetry and system events.
- 07
Closeout
Resolve failures or uncertainty, verify completion conditions and generate the full audit record.
Shared job model
Every participant works from the same bounded plan.
Worker-specific interfaces change. The job, task, permission, evidence and audit definitions do not.
Replaceable workers
Adapters connect execution without redefining the product.
A software agent, mobile human interface or machine adapter can accept the task shape it supports. Each still receives a bounded assignment and remains subject to the same state, policy and evidence rules.
The current public demonstration uses a simulated rover. It is one fictional adapter, not the Kinemica product category and not real machine-control validation.