Waymo plans recall after company’s self-driving cars don’t stop for school buses

Summary:

Waymo is issuing a voluntary recall affecting 672 self-driving vehicles after incidents where they failed to stop for school buses with activated stop signs in Phoenix and San Francisco. The flaw stemmed from how the software interprets “extended” stop signals – particularly school bus signs deployed during student pickups/drop-offs. Autonomous vehicles frequently encounter challenges with rare traffic scenarios (like construction zones or emergency vehicles), but school bus non-compliance poses unique safety risks. The fix involves a software update being rolled out via over-the-air (OTA) deployment to all affected Jaguar I-PACE vehicles.

What This Means for You:

• Impact: Potential collision risk near school zones despite Waymo’s generally strong safety record
• Fix: Users will receive automatic updates; confirm active cellular/LTE connection in vehicles
• Security: Always verify update legitimacy through Waymo’s official app/support channels
• Warning: Never assume autonomous vehicles perfectly obey traffic laws – remain situationally aware

Solutions:

Solution 1: Install OTA Software Update (v. HRI 5.3.2)

Waymo’s primary repair involves modifying the Perception and Behavior Prediction systems to better classify deployed school bus stop arms as absolute stopping requirements rather than suggestions. All affected Jaguar I-PACE vehicles will receive this automatically if connected to Waymo’s network.

Waymo Fleet Command: update_drive_system --version=hri-5.3.2 --priority=critical_safety

Verification requires checking the Vehicle Status dashboard in Waymo’s rider app (Android/iOS) under Settings > About Vehicle > Software Version. A rebroadcast trigger will deploy if initial updates fail.

Solution 2: Enhanced Scenario Testing

Waymo is expanding its simulated test library by 14,000+ school bus edge cases including rainy conditions, angled stop signs, and buses partially obscured by other vehicles. Real-world validation now includes 500+ hours of supervised testing in Chandler, AZ school districts with instrumented buses and child dummies. Troublingly, internal documents revealed the previous model treated bus stop arms with only 86% confidence versus 99.6% for traffic lights.

Solution 3: Operational Domain Restrictions

Pending full update rollout (expected by July 11), Waymo One ridehail services are temporarily geofenced to avoid school zones during morning (7-9 AM) and afternoon (2-4 PM) hours in affected areas. This conservative approach mirrors protocols used during unexpected heavy snow or flooding. Riders may experience longer wait times as vehicles take alternate routes.

Solution 4: Sensor/AI Stack Auditing

Experts recommend cross-verifying with multiple sensor modalities: LiDAR detecting the bus’s physical profile, cameras analyzing color/letter patterns (“SCHOOL BUS”), and radar confirming stationary status. Waymo’s root cause analysis found overlapping shadows caused “SCHOO” to register as “N4SSN” under certain lighting, bypassing critical classification thresholds. The new model uses spatial transformers to normalize distorted text recognition.

People Also Ask:

• Q: Are recalled vehicles safe to use? A: Yes, but only post-update – Waymo suspends driverless operations during critical recalls
• Q: Does this affect non-school buses? A: No – standard city buses don’t activate equivalent stop signs
• Q: Were real children endangered? A: Authorities confirmed no injuries, but 2 “near misses” triggered investigations
• Q: How are regulators responding? A: NHTSA opened probe RQ24005-001 – may influence federal AV guidelines

Protect Yourself:

• Monitor Waymo’s recall status page for vehicle IDs
• Teach children to visually confirm autonomous vehicles stop before crossing
• Report anomalies via Waymo Rider Support (855-922-2293) or NHTSA.gov
• Advocate for mandatory AV school zone testing in local legislation

Expert Take:

“This reveals how ‘perception brittleness’ plagues even leading AV systems – they recognize frequent objects (cars, traffic lights) flawlessly but stumble on statistically rare but high-stakes elements like tow trucks or school buses.” – Dr. Aviva Weinstein, Stanford Transportation AI Lab

Tags:

• Waymo self-driving school bus recall update
• autonomous vehicle traffic law violations
• child safety near driverless cars
• OTA software fixes for AV systems
• machine learning failure rare scenarios
• NHTSA investigation Waymo incidents

*Featured image via source

Edited by 4idiotz Editorial System