OpenForge: Collaborative Open Hardware Design & Manufacturing Platform

OpenForge is a cloud-based collaborative environment for open source hardware development, akin to GitHub but purpose-built for electronics and mechanical designs. It bridges the gap between design and production by offering integrated version control for CAD and EDA files, real-time collaboration (like Google Docs for schematics), cloud-based simulation and testing, and direct links to open hardware manufacturers for one-click ordering of PCBs and 3D prints. The platform is entirely open source, encouraging contributions and forks. Users can create projects, push design files (KiCad, FreeCAD, etc.), track changes, resolve conflicting traces, and spin up virtual test benches—all from the browser. By lowering friction for hardware collaboration, OpenForge aims to accelerate innovation in the open hardware ecosystem.

Open hardware projects suffer from fragmentation. Designers use disparate tools (KiCad, Eagle, FreeCAD) that don't talk to each other, collaboration is limited to emailing zip files or using clunky SVN, and manufacturing requires manual BOM building and sourcing. There's no unified, version-controlled, collaborative platform that handles the entire lifecycle from idea to assembled board. This means reproducible builds are rare, onboarding new contributors is hard, and many promising projects never make it to production.

• Provide git-based version control for hardware design files (schematics, PCB layouts, 3D models) with native diff/merge for common formats
• Enable real-time co-editing (both web-based and plugin-based) of design files
• Integrate cloud simulation for simple circuits and mechanical assemblies (e.g., SPICE, kinematic simulation)
• Automatically generate bill of materials (BOM) and integrate with open source friendly manufacturers (e.g., JLCPCB, OSHPark, PCBWay) for ordering
• Offer community features: project forking, issues, pull requests, discussions, and attribution (licensing like CERN OHL)
• Provide a web-based viewer for 3D models and PCB layouts without requiring local installations

• Not a replacement for professional-grade simulation tools (e.g., Ansys, SolidWorks) – we focus on lightweight validation
• Not a marketplace for selling hardware – strictly a collaboration platform
• Not a manufacturing fulfillment service – we connect to existing services
• Not aiming to replace KiCad or FreeCAD – we integrate with them via plugins
• Not supporting proprietary file formats – only open standard formats (or well-documented ones)

• Frontend: React with TypeScript, Three.js for 3D rendering, Monaco Editor for code/text views, WebSocket client
• Backend: Node.js with Express, PostgreSQL (metadata, users), Redis (caching, session), RabbitMQ (async jobs)
• File storage: Git-based repository service (custom built on top of libgit2 or isomorphic-git) stored in object storage (S3-compatible)
• Real-time collaboration: WebSocket server (ws library) with operational transforms for 2D/3D data
• Simulation: Docker containers running KiCad's Eeschema, FreeCAD's FEM, ngspice, etc., orchestrated via Kubernetes
• CI/CD: GitHub Actions, self-hosted runners for simulation jobs
• Security: HTTPS, JWT authentication, role-based access control for projects

Microservices architecture. API Gateway routes requests to services: Auth, Project, Repository, Collaboration, Simulation, Supply Chain. The Repository service manages git-like storage and diff/merge logic. The Collaboration service maintains WebSocket connections and merges real-time edits. The Simulation service spins up ephemeral Docker containers for user-initiated simulations. The Supply Chain service scrapes APIs of manufacturers for pricing and feeds BOM data. All services are deployable via Docker Compose or Kubernetes. Frontend is a single-page app communicating via REST and WebSockets. A plugin SDK lets existing EDA tools (KiCad, FreeCAD) push/pull changes directly from the application.

• Complexity of merging hardware files (non-text formats) – requires custom diff/merge algorithms
• Licensing concerns – hardware licenses (like CERN OHL) are less standardized; potential legal friction
• Scaling simulation containers could become expensive; need to limit compute per user
• Supply chain integration reliability – manufacturers' APIs change or break; handling errors gracefully
• User adoption – competing with GitHub's generic repos and native EDA tool plugins

• How to handle real-time collaboration on complex 3D models (e.g., overlapping edits)? Need robust conflict resolution strategy.
• Should we support non-CAD files (e.g., firmware, documentation) as part of the same repo? Probably yes, but that dilutes focus.
• What's the business model to keep servers running? Could offer paid tiers for private projects and extra simulation credits.
• How to ensure the BOM is accurate and up-to-date with component availability? Maybe rely on community maintainers.

The project never got past a detailed spec and prototype because the team got pulled into separate paid work and the sheer engineering effort required to integrate multiple EDA tools properly was daunting. It remains a 'what if' idea that still excites some of us.

Inspiration from GitLab for hardware. The plugin approach would be key – start with KiCad support, then FreeCAD. Community involvement could help implement diff/merge for other formats.