01
Capture provenance
Raw capture evidence, route context, timestamps, poses, device metadata, freshness, and source identifiers.
Proof
See what supports the site data and policy runs.
Blueprint proof keeps the buyer question grounded: what came from capture, what is inferred, what is still missing, and what claim the evidence can actually support.
Sample vs request
Public samples teach the workflow. Request packets prove one site.
Layer
Public sample
Request-specific proof
Site data product
Illustrative data-package support artifacts, scenario data, export manifest, and proof-boundary shape.
Capture record, task scope, robot profile, thresholds, rights/privacy posture, missing annotations, and proof gaps for one site.
Generated or model-derived output
Useful support signal when labeled as generated, inferred, or representative.
Owner-system evidence such as provider artifacts, runtime records, simulator traces, action logs, robot trials, and safety review.
Policy evaluation output
Shows how Blueprint frames proceed, modify site, gather more data, compare releases, or hold.
A request-scoped advisory tied to the exact facility, robot policy, pass bar, package state, and missing-proof labels.
Evidence hierarchy
Every output should point back to the source packet.
Repo tests and samples can prove public wording. They do not prove operational robot readiness.
02
Rights and privacy posture
Use limits, restricted zones, redaction needs, sharing scope, export limits, and commercialization boundaries.
03
Site data package
Manifest, included files, scenario data, linked geometry, proof depth, export limits, and package access review state.
04
Policy evaluation
Task thresholds, scenario variations, policy-run outputs, failure modes, annotation backlog, unresolved proof, and next-step recommendation.
A proof packet should make the data and policy decision obvious.
The robot team should be able to see why Blueprint recommends a policy session, more site data, site modification, release comparison, or hold.
Proceed to a short-pilot protocol
Modify the site before pilot time
Gather simulator traces, action logs, or robot-trial evidence
Compare vendors against the same task and threshold set
Hold until the proof gap is resolved
Claim boundary
Advisory until stronger proof exists.
Blueprint can publish polished site data, scenario, policy-run structure, sample packets, and request paths. It must not claim safety validation. Do not claim simulator execution completed, robot-trial success, or cleared rights. Do not claim hosted-session fulfillment, payment success, or ready to deploy unless the owning systems prove those facts for the request.
Request a proof packet