Waymo CEO on Reliability: Bridging the Gap from Demo to Product in 15 Years
机器之心 · wechat · 2026-08-15
Waymo Co-CEO Dmitri Dolgov shared the 15-year journey from early demos to scaled operations, highlighting the "reliability gap" in physical AI systems.
- Demo vs. Product: While the team verified core driving capabilities in just 18 months, it took another 15 years to build a deployable, reliable service.
- Physical AI Constraints: Unlike digital AI, physical systems face higher error costs, millisecond-scale decision requirements, data scarcity, and stringent pre-deployment validation needs.
- Reliability Scaling: Moving from 90% to 99.9% reliability increases difficulty exponentially, necessitating solutions for hardware aging, sensor failures, and long-tail events.
- Technical Strategy: The chosen tech stack must be evaluated not just on initial performance, but on its upper limit for reliability to support fleet scaling.
More from Embodied
- Meta Smart Glasses Compress Hardware into Ordinary Frames, Signifying Wearable Revolution — CurieuxExplorer · 2026-08-15
- Fall Detection System Using Pose Estimation and Deep Learning — rsasaki0109 · 2026-08-15
- New Framework Enables Multimodal Control for Humanoid Robots via Video and Music — zhengyiluo · 2026-08-15
- Humanoid Robots Deployed in Shenzhen Logistics Factory — rohanpaul_ai · 2026-08-15
- CyberOrigin Unveils Ground Truth Dataset for Embodied Intelligence with 123k+ Hours Captured — CyberRobooo · 2026-08-15
- From ESP32 to a Self-Perceiving AI: Open-Source Guide Builds a Body for Your AI — repligate · 2026-08-15