The Proof Layer

Building Trust

The hardest problem: How do we know a robot actually did the job?

Volm Network deploys the following techniques:

  • Sensor Proofs: The robot signs sensor outputs (e.g., GPS logs, video hashes).

  • Zero-Knowledge Proofs (ZKPs): Instead of uploading raw video (a huge bandwidth/privacy risk), the robot generates a ZK proof that it processed footage correctly.

    • Example: “I detected 15 objects in this video stream” → the proof verifies the computation was correct without exposing the video.

  • Remote Auditing: Third-party verifiers randomly challenge robots to perform micro-tasks (e.g., move the camera, return sensor data).

  • Hardware Attestation: We use ARM TrustZone/TPM to ensure the Volm Agent is running in a secure enclave.

This verification layer is critical to prevent fake participation (e.g., a malicious actor pretending to run drones).

Last updated