Building Kiteshield: A journey from prototype to safety-critical
summary
A Rust-based system proves location with ultra-wideband anchors mounted on molded cable segments to create safety bubbles that guarantee autonomous and manual vehicles never occupy the same space; the team rejects statistical validation because it would require 114,000 years of running, so it builds a formal mathematical proof down to measurement assumptions. Prototype infrastructure (running in multiple mines) uses tags and anchors at scale targets up to 100,000 anchors over 1,000 km of tunnels, integrates with vehicle e-stops, and demonstrates improved utilization during demos.
AI-written summary · about
chapters (16)
- 0:00Kiteshield project intro
- 1:05Need for mine automation
- 3:11Thrombosis problem in mining
- 5:41Suppliers failed on safety
- 7:04Founding Kiteshield initiative
- 8:29Functional safety approach
- 10:56Safety and open standards
- 11:38Demonstrating safety bubbles
- 14:09Productivity gains in Raviden
- 17:56Research to advanced engineering
- 18:59Kiteshield infrastructure
- 21:02Radio tech and deployment
- 23:29System architecture and scale
- 25:13Prototype in Rust
- 27:45Current status and plans
- 28:30Q&A
related talks
25mContinuous compliance with Rust in automotive software
37m"Rusty New 'ars: Look, Ma, No C Added!"
20m"Rust in Motion: Building Reliable and Performant Robotics Systems"
46mBuilding Real-Time Control Software: Lessons from Hyperloop
41m"Rust for Robotics: Safer, Faster Systems for Autonomous Applications"
56mRust for Foundation Software
27mDeterministic fleet management for autonomous mobile robots using Rust
45mBuilding Safe and Secure Systems in Rust