Yazılar

US Auto Safety Agency Reviews Tesla’s Responses on Robotaxi Safety Ahead of Deployment

/in /tarafından

The U.S. National Highway Traffic Safety Administration (NHTSA) is currently reviewing Tesla’s answers regarding the safety of its upcoming robotaxi service, particularly how the self-driving vehicles will perform in poor weather conditions. This review comes ahead of Tesla’s planned limited launch of its robotaxi service in Austin, Texas, which could begin as early as this weekend.

Tesla has reportedly invited a small group of people to participate in a test of the robotaxi service, with posts and screenshots on social media showing that a Tesla employee will accompany riders in the front passenger seat. NHTSA had requested detailed information from Tesla by June 19 to better understand the company’s safety protocols and technology for the robotaxi rollout.

The agency’s ongoing probe covers 2.4 million Tesla vehicles equipped with Full Self-Driving (FSD) technology, following multiple collisions in reduced visibility conditions, including a fatal crash in 2023. NHTSA’s questions focus on how Tesla plans to ensure safety in challenging conditions such as fog, sun glare, rain, dust, or snow, and what the vehicle’s response would be if such conditions arise during a ride.

Additionally, the agency has sought clarity on how many vehicles will be deployed as robotaxis and the timeline for wider availability, especially for vehicles operated by third parties rather than Tesla directly. Tesla CEO Elon Musk has emphasized a strong safety focus for the upcoming trial, with humans remotely monitoring the vehicles during operations.

Tesla did not immediately respond to requests for comment, and NHTSA said it will update the public file once its review is complete.

Tesla to Use Remote Teleoperation for Robotaxis Amid Safety Concerns, Faces Limitations

/in /tarafından

Tesla plans to launch its robotaxi service in Austin, Texas, with around 10 Model Y SUVs operating under strict safety limits. The vehicles will be remotely monitored by humans through teleoperation, a technology allowing operators to control cars from a distance via wireless networks. Tesla CEO Elon Musk emphasized a cautious approach, with humans ready to take control in challenging situations, such as navigating crowded areas or confusing traffic scenarios.

Teleoperation in the robotaxi industry varies. For example, Waymo’s vehicles mostly operate autonomously but can consult human “fleet response” agents for help when uncertain. Baidu’s Apollo Go offers fully remote backup drivers who can take over virtually. Tesla’s model leans toward remote control intervention, supported by recent job ads seeking teleoperation personnel capable of managing vehicles and robots remotely.

However, teleoperation has key limitations. It depends on stable cellular connections, which can lag or drop, creating safety risks if vehicles lose contact with operators during critical moments. Experts warn that while teleoperation might be feasible for a small test fleet like Tesla’s, it becomes unreliable at scale. Additionally, one operator can only monitor a limited number of vehicles effectively.

Tesla’s robotaxis will initially operate only in safe, limited parts of Austin, avoiding complex intersections. The company’s Full Self-Driving software will provide autonomous control, with teleoperators acting as a safety backup. Some Texas lawmakers have urged Tesla to postpone the launch until new autonomous driving laws take effect in September to ensure public safety and trust.

Honda-backed Helm.ai Unveils Vision System for Self-Driving Cars

/in /tarafından

Helm.ai, a California-based startup backed by Honda Motor, introduced its new camera-based urban environment interpretation system called Helm.ai Vision. The company is negotiating with multiple automakers to integrate its self-driving technology into mass-market vehicles.

Helm.ai is collaborating with Honda to embed the system in the upcoming 2026 Honda Zero series of electric vehicles, which will enable hands-free driving and allow drivers to take their eyes off the road.

CEO and founder Vladislav Voroninski told Reuters that the company’s business model centers on licensing this software, including foundation model software, to automakers. Helm.ai’s vision-focused system aligns with Tesla’s approach, relying on cameras rather than sensors like lidar or radar, which can add significant costs.

Voroninski acknowledged Helm.ai’s foundation models can work with other sensors but emphasized that the primary offering remains vision-centric. Industry experts, however, highlight that supplementary sensors such as lidar and radar are vital for safety, especially under poor visibility conditions.

In contrast, robotaxi companies like Alphabet’s Waymo and May Mobility use a sensor fusion approach combining radar, lidar, and cameras to ensure comprehensive environment perception.

Helm.ai has raised $102 million to date, with investors including Goodyear Ventures, Korean auto parts maker Sungwoo HiTech, and Amplo.

The Helm.ai Vision system merges inputs from multiple cameras to create a bird’s-eye view map that enhances vehicle planning and control. It is optimized for hardware platforms from Nvidia, Qualcomm, and others, facilitating automakers’ integration of Helm.ai Vision into existing vehicle systems.