Tenzro
Reference design — PhysiClaw

Verifiable physical AI on Tenzro.

A reference design applying TenzroClaw to physical AI systems — robots, drones, and IoT devices that interact with the physical world using verifiable execution, attested sensors, and trustless settlement.
Overview

What PhysiClaw looks like, layered onto Tenzro Network.

PhysiClaw is a hypothetical extension of the OpenClaw skill stack into physical embodiments. Each device gets a TDIP autonomous-class DID with auto-provisioned wallet. Sensor data is attested through TEE-sealed channels. Consequential actions emit Plonky3 STARK proofs over the perception-action loop. Settlement for goods delivered or services performed runs through Tenzro's payment and settlement engines.
Architecture

Primitives PhysiClaw would use.

Autonomous device DIDs

Each robot or drone gets did:tenzro:machine:{uuid}. Stakes TNZO. Owns its wallet. Holds its reputation.

TEE-sealed sensors

Sensor streams attested through Intel TDX, AMD SEV-SNP, AWS Nitro, or NVIDIA GPU CC. Tampering detectable cryptographically.

ZK action proofs

Plonky3 STARKs over perception-action transitions. Verifiable that a robot saw X and acted Y.

AgentBond + insurance

Devices post bond. Counterparties file insurance claims for damages. Network resolves disputes.

Multi-modal inference

Vision encoders (CLIP, SigLIP2, DINOv3), segmentation (SAM 2), detection (RF-DETR, D-FINE) — onboard or routed.

Settlement on delivery

Escrow released on attested completion. Channel updates for streaming services. Cross-chain payouts via bridge router.

User journey

From device provisioning to verifiable service.

Get started

Ship on the open network.