Hive × Physical AI · the vertical

Provable Machines

Wheeled, legged, flying, humanoid. A new generation of machines is heading for factory floors, warehouses, streets, skies, and living rooms, and every one of them now navigates and acts on its own from a camera and a plain-language instruction. The first serious incident will come down to one question: what did the machine see, decide, and do, and when? The manufacturer's own telemetry is a document written by the defendant. Hive gives every machine an independent, signed, timestamped witness that works the same way no matter what the machine looks like.

1M / yr
production targets already announced across the sector
wheeled · legged · flying
one navigation brain, many machine types, one record format for all of them
~$30 / hr
robots-as-a-service: metered work needs metered proof
Art. 12
EU AI Act + Machinery Regulation: logging is now law, not hygiene

Sidecar to the control loop. Never in it.

The witness chain · animated

Sense → Perceive → Decide → Actuate. Signed at every hop.

AFiR-Stream anchors the session. Each stage emits receipts on its own while the control loop keeps running untouched at full speed. One receipt binds the camera frame, the plain-language instruction, the decision it made (a waypoint, a pointing target, a move), and which machine carried it out. That means one record format works the same across wheeled, legged, flying, and humanoid machines. When nothing goes wrong, the receipts are a compliance asset you can check offline against the Hive public key, the same way in every vertical. When something does go wrong, press the red button.

The clock is real. The EU AI Act's high-risk logging and record-keeping duties (Article 12) phase in through Aug 2, 2026 and Aug 2, 2027, and the EU Machinery Regulation (2023/1230) applies from Jan 20, 2027. Both make a tamper-proof record of what an autonomous machine saw and decided a legal requirement, not a nice-to-have. Any machine shipped into the EU is in scope. An independent, timestamped receipt is how you meet that requirement without handing a regulator your raw sensor data or your model weights.

Sense
RGB camera · lidar · force-torque
frame hashes, not frames
SiGR
Perceive
model + version bound
world-state attested
AFiR™
Decide
instruction → waypoint / action
plan · envelope · receipted
AFiR-Exec
Actuate
command issued · motion
session-anchored stream
AFiR-Stream
control loop: untouched · full rate · receipts ride sidecar, ~70ms, async

The red button is not a kill switch. It runs the evidence · export · escalate flow. It never touches the control loop.

The software agent is a machine too

An agent that ships its own fix runs the same loop, and needs the same receipt.

A robot senses, decides, and acts in the physical world. An autonomous ops or coding agent does the same thing in a codebase: an alert fires, it reads the logs, it figures out a root cause, and it acts, opening a pull request, shipping a fix, restarting a service. The new wave of incident-response agents wires straight from an alert to a merged change with no human in the loop. That's the same sense, decide, act chain, and it carries the same risk. When the auto-shipped fix takes down checkout at 3am, the only record of what the agent saw and why it acted is the agent's own session log, a document written by the thing under investigation.

Put the witness chain underneath the agent and every step gets signed the instant it happens: which alert fired, which logs it read, the root cause it claimed, and the exact change it pushed. It's bound by hash, anchored outside the agent's own boundary, and checkable offline. Sign-before-act means the receipt for the action exists before the pull request is opened, not pieced together after the outage.

Alert
Sentry · Datadog · Rootly
alert payload hashed, bound
SiGR
Investigate
logs read via MCP
model + version bound
AFiR™
Decide
root cause → planned change
scope · envelope · pre-cleared
Imprimatur
Ship
PR opened · fix merged
diff bound by hash
AFiR-Exec
the agent runs at full speed · receipts ride sidecar · sign-before-act on the Ship hop
Receipt one agent action · the live signer, no key, no wallet
# an incident-response agent about to open a PR: sign the action before it acts LOGS=$(sha256sum incident.log | cut -d' ' -f1) # hash the logs it read, never uploaded curl -sS -X POST https://receipts.thehiveryiq.com/v1/protected-flow/assess \ -H 'Content-Type: application/json' \ -d '{"policy_pack":"default","declared_exposure_usd":10.0, "receipts":[{"action":"agent_action","external_system":"incident-response", "actor":{"did":"did:example:ops-droid"}, "payload":{"stage":"decide","alert_source":"datadog", "logs_sha256":"'$LOGS'", "root_cause":"duplicate backend requests from retry storm", "action":{"type":"open_pr","repo":"api"},"event":"auto_fix"}}]}'

Returns a live signed assessment with a verifiable group_id, the same production signer the robots use, unchanged. Autonomy is the easy part now. Provable autonomy, a signed record a regulated on-call team can actually sign off on, is the part the leaderboards skip.

Run it like the robot · live, free, no wallet

Give the machine a decision. Get a signed receipt back.

This is the real production signer, not a mock. Send one navigation decision (the plain-language instruction, a hash of the camera frame it acted on, and the action it chose, a pointing waypoint, a heading, a move) and get back a live Ed25519 signed receipt bound to the exact bytes, checkable offline by anyone. Wrap it around a control loop and every step the robot takes produces its own receipt: a signed, timestamped record of what it saw, what it was told, and what it decided. Video is bound in by frame hash, so no raw video ever leaves the machine.

# press “Sign a robot decision” and a live Ed25519 receipt for one navigation step appears here # instruction → “leave the lobby, walk through the corridor, stop to face the second shelf” # decision → pointing waypoint [0.62, 0.48], heading 25° · frame bound by SHA-256, never uploaded
Grab the curl · one decision, one receipt
# copy · paste · run: free tier, no key, no wallet FRAME=$(sha256sum frame.jpg | cut -d' ' -f1) # hash the RGB frame the robot acted on curl -sS -X POST https://receipts.thehiveryiq.com/v1/protected-flow/assess \ -H 'Content-Type: application/json' \ -d '{"policy_pack":"default","declared_exposure_usd":10.0, "receipts":[{"action":"machine_actuation","external_system":"machines", "actor":{"did":"did:example:mobile-unit"}, "payload":{"stage":"decide","frame_sha256":"'$FRAME'", "instruction":"proceed to the second shelf", "decision":{"waypoint":[0.62,0.48],"heading_deg":25},"event":"nav_step"}}]}'
Wrap the control loop · a receipt per step
# pseudo-loop: every decision the robot makes gets its own signed receipt for step in robot.run(instruction): frame_hash = sha256(step.rgb_frame) # audiovisual, bound not uploaded r = POST("/v1/protected-flow/assess", { "receipts": [{ "payload": { "frame_sha256": frame_hash, "instruction": step.instruction, "decision": step.action } }] }) receipts.append(r.signed_assessment) # Ed25519, verifiable offline # → a signed timeline of every movement, decision, and frame the run touched

The open demo runs on the free assessment tier and returns a signed decision plus a verifiable group_id. The paid production path (/v1/receipt/emit, per-decision mint, and the audiovisual Media Origin Receipt) adds the full post-quantum envelope, ML-DSA-65 plus SLH-DSA, settled per receipt over x402. Checking a receipt is always free and works offline against the Hive public key.

Proof state · receipt → vector → signed decision

From a pile of stage receipts to one signed decision

The witness chain signs each hop. R3Pv™ groups those receipts into a single proof vector: verification depth, the honest recoverability window, and the actions allowed next, all in one object. Protected Flow is the action layer above it. One call turns a machine event into a signed decision (permit · permit_with_evidence · require_approval · hold · recover · reroute · escalate · block) plus a metered quote and an evidence bundle you can export. That's exactly what the red button fires.

R3Pv signed proof vector · schema r3pv-v1.0.0
// derived from the grouped stage receipts of one machine session { "signed_vector": { "vector_id": "r3pv.vector_…", "key_id": "did:hive:hivemorph", "algorithm": "Ed25519", "vector": { "schema": "r3pv-v1.0.0", "weakest_proof_boundary": "self_attested", "healing_state": "evidence_only_final", "routing_recommendation": "require_approval", "permitted_next_actions": ["hold","export_evidence"] } }, "verification": { "ok": true, "verify_endpoint": "/v1/receipt/verify" } }
Protected Flow signed decision · schema protected-flow-v1.0.0
// one call: POST /v1/protected-flow/assess, real production response { "signed_assessment": { "key_id": "did:hive:hivemorph", "algorithm": "Ed25519", "assessment": { "schema": "protected-flow-v1.0.0", "decision": "hold", "permitted_next_actions": ["hold","export_evidence","preserve_evidence"], "policy_pack": "default", "meter_quote": { "total_usd": 1.105, "note": "quote, no funds move" }, "group_id": "r3pv.group_…" } }, "verification": { "ok": true, "verify_endpoint": "/v1/receipt/verify" } }

A machine event assessed at the self_attested boundary honestly returns hold. The default pack requires relay_observed before it will permit, and once a physical incident is final the vector reports evidence_only_final. The receipt never claims a finished action can be undone. Policy can only make a decision stricter. The meter is a quote, not a settlement. Both signed objects round trip through POST /v1/receipt/verify.

# the red button, on the live rail: one call to assess, then export the evidence bundle GID=$(curl -sS -X POST https://receipts.thehiveryiq.com/v1/protected-flow/assess \ -H 'Content-Type: application/json' \ -d '{"policy_pack":"default","declared_exposure_usd":25.0, "receipts":[{"action":"machine_actuation","external_system":"machines", "actor":{"did":"did:example:mobile-unit"}, "payload":{"stage":"decide","instruction":"proceed to the second shelf", "decision":{"waypoint":[0.62,0.48],"heading_deg":25},"event":"nav_step"}}]}' \ | python3 -c 'import sys,json;print(json.load(sys.stdin)["signed_assessment"]["assessment"]["group_id"])') # export the signed evidence bundle for the deployer, insurer, or examiner curl -s -w '\nexport %{time_total}s %{http_code}\n' \ "https://receipts.thehiveryiq.com/v1/protected-flow/evidence/$GID/export"
● Smoke tested · 2026-07-07 · production

GET /v1/protected-flow/health and POST /v1/protected-flow/assess returned HTTP 200 (signed decision hold, meter quote 1.105). The assessment verified verified=true at /v1/receipt/verify, and /v1/protected-flow/evidence/{group_id}/export returned a 200 signed bundle. Observed round trip: 0.11 to 0.30 seconds including network. Measure receipt latency separately from control-loop or rail latency. Measured latency & runnable curl →

The question every household asks

Who's driving? The model, or a person?

Home humanoids are trained and assisted by human teleoperators. That's fine, until nobody can prove when. Human-Origin Attestation stamps every session with who or what was in control, without exposing a single camera frame. It's the same primitive Hive uses to prove human versus agent origin across every AI-authored action. Flip it around.

Autonomous
Teleop

CLOAzK™ proves the behavior without exposing the home. Warden receipts the deletion. HiveSeal · Colony silicon roots ensure the receipt itself can be trusted.

The coverage question · answered

Every hard question a robot raises. A primitive for each.

Scroll down. The checks land one by one.

✓ COVERED
What did the fleet's model do before last night's OTA update, and can you prove today's robot is different?
Agent Metamorphosis Attestation · SiGR
✓ COVERED
Was the safety envelope checked before the arm moved near a person?
AFiR-Exec · Safety-Envelope Attestation
✓ COVERED
Two robots coordinated a lift. Who decided what?
Agent Coalition Attestation
✓ COVERED
Can the receipt itself be trusted if the robot's software stack is compromised?
HiveSeal™ / Colony: signed at the silicon, USB HSM root
✓ COVERED
Prove the home robot behaved, without handing over bedroom camera footage.
CLOAzK™ zero-knowledge posture · Warden disposition receipts
✓ COVERED
Bill 40,000 RaaS hours across three customers, with proof for every metered hour.
SLS Metering · x402 settlement rail · Settle
✓ COVERED
Ship the same fleet into the EU, Texas, and Japan, and prove which rules applied to each unit.
Jurisdictional Clearance
✓ COVERED
A human took over mid-task. Exactly when did control transfer?
✓ COVERED
The incident is final. Nothing can un-happen. What now?
R3Pv™: the receipt never claims more recoverability than it has. Evidence only, honestly.
9 / 9 · full-stack coverage · all primitives patent pending
The insurer's dashboard · configure a fleet

Underwriters don't insure promises

Nobody writes coverage on a 10,000-unit humanoid fleet backed by self-kept logs. Set the fleet; watch the posture gauge move from "uninsurable narrative" to "provable machine." This dashboard is the conversation with the carrier.

Fleet size (units deployed)500
Operating hours per unit / day16
% of consequential actions receipted40%
Hardware root (HiveSeal™ units)0%
0 /sec
Receipts signed
0 M/mo
Evidence volume
Not set yet
Underwritability
Three buyers · one rail

Who signs the check

The Maker

Shipping thousands of units into other people's buildings. Every unit is a liability node until it can testify for itself. Provable-by-design becomes the spec line competitors don't have.

Hook: the first maker to ship receipts sets the standard the rest must follow.

The Deployer

Warehouses and factories running mixed fleets they didn't build. When the incident review comes, "the vendor's logs say" is not a position. Independent receipts are the deployer's leverage over every vendor.

Hook: one rail across all makes. The deployer's evidence doesn't depend on the maker's cooperation.

The Insurer

The de facto regulator of physical AI. Premiums price uncertainty. Receipts remove it. A "provable machine" class of risk is a product a carrier can actually underwrite, and discount.

Hook: the carrier that pilots receipt-based pricing owns the category before actuaries elsewhere wake up.

The quiet pilot · one robot cell · three weeks

Shadow-receipt a single work cell or a ten-unit fleet. Zero changes to the control stack. The sidecar listens, anchors, and signs. It ends with one Deployment Evidence Bundle: perception-to-action chains, teleop attestations, and OTA history, all checkable offline against the Hive public key. The pilot's real product is the document your safety lead forwards.

Same primitives · every vertical

Provable Machines is one Hive vertical. Here's the rest.

The receipt shape that anchors a humanoid's control loop is the same one that anchors a sportsbook's intervention, a bank's dispute, or an AI agent's payment. One primitive set. Many verticals. All verifiable offline against the Hive public key.

Primitive index · click through
AFiR™ AFiR-Stream AFiR-Exec SiGR HiveSeal™ Colony R3Pv™ Protected Flow Human-Origin Attestation Cryptographic AI Receipts Settle x402 Checkout Wallet CLOAzK™ · Warden · Metamorphosis · Safety-Envelope · Coalition · Jurisdictional Clearance · SLS Metering: patent pending
The live rail: verify it now
# the same rail every vertical above uses curl https://receipts.thehiveryiq.com/v1/receipt-relay/health # -> {"status":"ok","signer":"did:hive:hivemorph","algo":"Ed25519"}
Private workspace · not affiliated with any robot manufacturer named or implied · no manufacturer data processed · all demo endpoints are Hive's own · all primitives patent pending · live rail: receipts.thehiveryiq.com
Private by design. Hive does not store your prompts. Every request is already receipted by a one-way SHA-256 fingerprint, not the words. Proof, not surveillance.