Success rate
Can the robot complete the task often enough?
The evaluation starts with the pass bar the robot team actually needs.
Real-site robot evaluation
Evaluate robots on real sites before deployment.
Blueprint turns captured facilities into robot task packs so teams can test policies, find failure modes, and prepare for shorter field pilots.
Robot teams pay for evaluations. Site operators can submit sites free.
Task Evaluation Run
One Task Evaluation Run = 1 site × 1 robot policy/profile × 1 Task Pack × up to 500 scenarios.
Cycle time
Can it keep up with the site rhythm?
Blueprint frames the task against target timing, bottlenecks, and site drift.
Intervention rate
Where will people still need to step in?
Blueprint keeps assist points visible instead of hiding them behind a score.
Safety threshold
What still needs review before field exposure?
Safety stays scoped to the request and does not become a blanket validation claim.
What Blueprint sells
Two robot-team products.
Start with a Task Evaluation Run when you need a scoped answer before field time. Add a Policy Improvement Run when the team wants Blueprint to improve a supplied policy inside simulation.
01
Task Evaluation Run
One site, one robot policy/profile, one Task Pack, and up to 500 scenarios.
02
Policy Improvement Run
Baseline eval, failure diagnosis, twin and cousin scenarios, sim-only curriculum, candidate policy improvement, sealed test, and evidence report.
How it works
Turn a real site into an evaluation plan.
Blueprint keeps the workflow compact: one site, one robot profile, one task pack, one integration mode, scored scenarios, and the next proof step.
Start with a real site
Use an existing site package, lawful capture, or structured facility request.
Describe the robot and task
Name the robot size, reach, sensors, controller level, task, thresholds, start states, and scenario variations.
Choose the integration mode
Use a policy API, container, private-cloud run, action trace, or site package only when the team keeps its stack in-house.
Run or review the scenarios
Measure success, cycle time, intervention points, safety events, and failure modes from simulator traces, action logs, or review evidence.
Decide the next proof step
Proceed to pilot, request more data, tune on the exported set, or hold until blockers clear.
Planning ranges
Two paid robot-team products. Site operators submit free.
Task Evaluation Runs start at $6,500 per run. Policy Improvement Runs start at $35,000 per sim-only run. Site operators can submit sites for free.
Task Evaluation Run
From $6,500 / run
Test one robot policy/profile against one real-site Task Pack, up to 500 scenarios.
Request Task Evaluation RunPolicy Improvement Run
From $35,000 / run
Improve a customer-supplied policy, adapter, task head, distilled skill, or full policy inside a sim-only run.
Request Policy ImprovementSite operators submit sites free.
Facility owners can submit or claim a site, define privacy/access boundaries, and review commercial-use terms before anything is shared.
Submit site freeEvidence boundary
Public examples show the workflow shape.
A request packet proves one site, robot profile, task pack, thresholds, rights posture, and missing proof. Evaluation output remains advisory until supported by simulator traces, action logs, robot trials, safety review, and runtime evidence.
First request
Have one real site or task to evaluate?
Bring one site, task, robot profile, and target threshold. We'll recommend the right evaluation scope, scenario count, and next proof step.