Risky changes hit hardware without rehearsal
Teams deploy complex changes directly to production without proper simulation testing, leading to unexpected failures and safety incidents.
Logs are debatable and incomplete
When incidents occur, teams struggle to determine what happened and why. Logs can be tampered with and lack context.
Cloud-only approvals fail at the edge
Robots operating in remote locations can't wait for cloud-based approval systems. Network failures prevent critical decisions.
Robots are moving from controlled environments to real-world deployments. AMRs in warehouses, delivery drones, and inspection robots are now operating at scale, creating demand for verifiable operations.
Enterprise customers and regulatory bodies increasingly require proof of safety and compliance. Insurance companies demand verifiable risk assessments, and regulators want audit trails for autonomous systems.
Edge computing capabilities and modern cryptographic primitives (Ed25519, SHA-256, Merkle trees) make it feasible to generate and verify proofs at the edge without compromising performance.
Sim-gated by default for risky changes
Edge quorums approve or halt under local constraints
Signed attestations bind who, what, when, why
Create and verify attestations (Ed25519, SHA-256), bundle telemetry with replay manifest, local static viewer for inspection.
Logs → reproducible scenarios, parameter sweeps at scale, Gate DSL for pass/fail, signed promotion or rollback decisions.
Short-lived quorums at the edge, t-of-n signatures from robots/operators/safety sources, clear local UI and time windows.
Content-addressed action bundles, periodic anchoring to public data root, light client verifies inclusion and signatures.
Every risky change should rehearse before it matters
SGO turns real logs into reproducible simulation scenarios, runs sweeps at scale, enforces policy gates and produces a signed decision you can trust. Faster rollouts, fewer incidents, verifiable outcomes.
Converts logs into deterministic sim scenarios
Runs parameter sweeps (speed, friction, environment) in parallel. Evaluates outcomes with a Gate DSL (collisions, time, map overlap, energy). Emits a signed decision (promote/rollback) plus HTML + JSON reports.
Safer releases, operational truth, audit-ready
Changes must prove themselves before touching machines. Outcomes tied to telemetry and policy context, not screenshots. Every decision comes with an attestation anyone can verify.
Pilot integrations and demo recording
Pilot integrations for fleets running AMRs, drones or robotic cells. Demo recording will be published to show SGO end-to-end in action.
ROS2 production for AMRs and inspection, repeatable sim from rosbags
DDS QoS, time sync, observability, on-site gateways
Signatures, content addressing, threshold flows
CLI ergonomics, HTML reports, CI pipelines
Collaboration with TheHumanoidAI
on brand and narrative development
Approached by multiple ecosystems
for funding and GTM support
Bottom-up adoption, top-down trust
Open-source entry point via Attest CLI, design partnerships with fleets and simulator vendors, proof-driven content marketing.
Simple unit logic
Land via open core → convert to hosted provenance → expand with policy packs and EQG seats. Target gross margins in line with modern SaaS.
Tight capability windows with expiry
Agents operate with minimal permissions by default, with time-limited access tokens that automatically expire.
High risk needs visible human presence
Critical operations require approval from multiple authorized personnel, with clear audit trails of who approved what and when.
Signed builds and replayable context
Sims and replays run in isolation with signed artifacts and SBOM checks. Every action is cryptographically verifiable.
Proofs without leaking sensitive data
Our attestation system provides cryptographic proof of actions without exposing proprietary algorithms or sensitive operational data.
To accelerate SGO pilots and EQG runtime development, expand our engineering team, and scale our go-to-market efforts.
Intros to fleets running 10–200 robots, simulator vendors, and enterprise customers who need verifiable robotics operations.
Help formalize proof standards with buyers, insurers, and regulatory bodies to establish PitlaneOps as the industry standard.