AI-driven triage, root cause analysis, and automated remediation for production ROS 2 environments. Deployed on your infrastructure.
20 min technical walkthrough - no commitment
The problem
Manual diagnostics don't scale. Your best engineers are stuck triaging instead of building. selfpatch closes the loop.
Engineers spend 2-4 hours per incident grepping logs, reproducing failures, and guessing root causes across distributed systems.
AI-guided diagnosis in under 4 minutes.
Only 1-2 senior engineers can diagnose complex incidents. When they're unavailable, MTTR doubles and SLAs are at risk.
Institutional knowledge, codified and repeatable.
Each robot down means missed throughput targets, delayed orders, or halted production lines. The cost compounds with fleet size.
Fleet-level triage: resolve what matters first.
How it works
From failure detection to verified resolution.
Continuous monitoring of runtime signals, fault events, and diagnostic messages across every node in your fleet.
AI correlates signals across robots to pinpoint root causes. Ranked hypotheses replace hours of manual log analysis.
Guided remediation with human-in-the-loop approval. Configuration fixes, parameter rollbacks, or service restarts - verified and traceable.
Platform
SOVD-compatible REST API for ROS 2. Discover nodes, read topics, trigger operations, manage configurations. Full source, Apache 2.0.
Production-grade diagnostics and self-healing for mission-critical fleets. AI agents, runbooks, fleet management, and dedicated support.
Security
Built for production environments with strict security and compliance requirements.
Deploy on-prem, in your VPC, or air-gapped. Diagnostic data never leaves your environment.
TLS encryption, JWT-based RBAC, audit logging. SOC 2 Type II preparation underway.
Dedicated support with guaranteed response times. Custom runbooks for your fleet's specific failure modes.
Full requirements traceability, ISO 26262 awareness, GDPR-compliant data handling. Audit-ready documentation.
Standards
ASAM Service-Oriented Vehicle Diagnostics
Jazzy LTS - production middleware
Model Context Protocol for AI agents
Open core, zero vendor lock-in
FAQ
Quick answers about selfpatch.ai
selfpatch.ai is a self-healing diagnostics platform for ROS 2 robot fleets. It helps engineering teams reduce Mean Time To Repair by detecting failures, diagnosing root causes with AI, and executing guided remediation.
A diagnostic agent runs on your robots and exposes a REST API. When incidents occur, AI analyzes patterns across your fleet, correlates events, and presents ranked hypotheses for your engineers to validate.
Currently ROS 2 Jazzy LTS. The agent runs as a standard ROS 2 node that integrates with your existing workspace. Humble support is on our roadmap.
From single-robot environments to fleets of 500+. Larger fleets benefit from better pattern detection and cross-robot incident correlation.
Edge agents run on the robot itself. Cloud/fleet components deploy to your infrastructure - AWS, GCP, Azure, or fully on-premises. Air-gapped deployments supported for sensitive environments.
Typical timeline: Week 1 - agent deployed and baseline established. Week 2 - first fault clusters identified. Weeks 3-6 - fleet-wide incident correlation fully operational.
We start with a paid pilot to demonstrate value on your specific fleet. Production pricing is based on fleet size and support tier. Contact us for a custom quote.
Diagnostics data stays in your environment at all times. Edge agents use TLS encryption and JWT-based RBAC. We are preparing for SOC 2 Type II certification. No diagnostic data is transmitted to external services.
The diagnostics gateway (ros2_medkit), web UI, and MCP adapter are all Apache 2.0 licensed. Enterprise features - AI agents, fleet analytics, protocol bridges, and SLA support - are commercial.
Have a question not answered here? contact@selfpatch.ai
Get started
20-minute call with a solutions engineer. We'll discuss your challenges and show you how selfpatch fits.
Prefer email? contact@selfpatch.ai