Solus
Grand Prize at VentureHacks
A robotics development workspace that builds a living knowledge graph across your entire robot system. Connects hardware schematics, software repos, CAD assemblies, and runtime telemetry into a unified model, then uses that understanding for automated change impact analysis, cross-domain debugging, and AI-powered root cause detection. Includes a real-time telemetry dashboard with anomaly detection, an in-browser MuJoCo simulator with PID auto-tuning, and intelligent issue search across historical fixes.
Solus is a robotics context model that builds a living knowledge graph of your entire robot system, spanning hardware schematics, software repositories, CAD assemblies, runtime telemetry, and team documentation. It understands the robot as a full system and uses that understanding to detect change impact, debug across domains, and surface reusable team knowledge.
Context Engine
The core of Solus is a graph database modeling 14 entity types (electrical parts, software modules, interfaces, runtime signals, ROS topics, and more) with typed relationships extracted from source integrations. Connectors ingest KiCad schematics, GitHub repos, Onshape assemblies, and PDF documentation, automatically discovering cross-domain relationships through static analysis.
When any component changes, BFS-based impact analysis traverses the graph to surface every downstream dependency, with AI-generated explanations for each affected node.
AI Agent
A 6-mode query router handles natural language questions about the system: debugging queries, architecture lookups, impact analysis, memory search, change history, and cross-domain tracing. The agent grounds every response in the knowledge graph with source pills and confidence scores.
Live Bench
A real-time telemetry dashboard streams simulated sensor data at 2Hz over WebSocket, rendering signal grids with sparkline charts and threshold-based anomaly detection. Historical issues are indexed with TF-IDF for similar issue search, linking runtime anomalies back to root causes in hardware or software.
MuJoCo Simulator
An in-browser MuJoCo WASM simulator renders the robot in 3D via Three.js, with a built-in PID auto-tuner that optimizes control parameters and shows before/after comparison against target trajectories.