Self-Driving & Robotics Quarterly Report — Q2 2026

Q2 2026 delivered three landmark commercial events that together define this quarter as the sector's most consequential since Waymo's San Francisco driverless launch in 2022. Aurora launched the first commercial driverless Class-8… 1. Executive Summary Q2 2026 delivered three landmark commercial events that together define this quarter as the sector's most consequential since Waymo's San Francisco driverless launch in 2022. Aurora launched the first commercial driverless Class-8 trucking service on the Dallas–Houston I-45 corridor in April, moving autonomous freight from a validated research project to a revenue-generating business. Figure AI's humanoid robot commenced supervised commercial assembly tasks at BMW's Spartanburg, South Carolina plant — the first large-scale industrial deployment of a general-purpose humanoid in an unmodified automotive production environment. And Waymo expanded commercial driverless service to Atlanta, Georgia — its first market outside California and Arizona — demonstrating that its operational model can transfer to new geographies. Against these advances, CVPR 2026 (June) and ICRA 2026 (May) surfaced a wave of end-to-end driving and foundation-model manipulation papers that are shortening the gap between academic capability and production deployment. The primary challenges this quarter were the public controversy over worker displacement in Aurora's trucking corridor, a reported near-miss incident involving a Waymo vehicle and an emergency responder in Atlanta, and ongoing uncertainty about the US Senate's path to a federal AV liability statute. --- 2. Technical Updates — Previous Quarter (Q1 2026) 2.1 Autonomy Stack & Perception Q1 2026 saw the first production deployment of NVIDIA Alpamayo-based compute in commercial AV development pipelines. Three Tier-1 OEM suppliers — Continental, Aptiv, and Bosch — confirmed Alpamayo integration into their next-generation ADAS platforms, citing the open-source driving vision-language models as the primary enabler for rapid customisation on their proprietary sensor configurations. Waymo deployed a camera-first perception layer to 35% of its San Francisco fleet in a shadow-mode trial — a significant architectural shift toward camera redundancy that reduces per-vehicle sensor cost by an estimated $4,200. Tesla FSD v13 shipped to approximately 820,000 enrolled vehicles in the US by the end of Q1, extending end-to-end neural control to adverse weather scenarios including light snow and heavy rain. Internal FSD intervention rate data (shared at the February earnings call) showed a 41% reduction in driver-requested disengagements per mile compared to v12.5, though Tesla did not release third-party verified mileage counts. Performance log (Q1 2026 actuals vs. Q2 2026 targets): Metric Q1 2026 (Actual) Q2 2026 (Target) Notes -------- ----------------- ----------------- ------- Waymo disengagement rate (per mile) 0.00188 0.00155 Driverless commercial fleet Miles between critical interv

More reports