In autonomous and self-driving vehicle news are NHTSA, Tesla, Waymo, Bot Auto, Lenovo, NVIDIA, PlusAI, Texas Ventures III, TomTom, DeepRoute.ai, Valeo, KE:SAI, TIER IV, Arm,Taisei, Nissan & Volvo Cars>
NHTSA, Tesla, Robotaxi School Zone Speeding Allegations Under Review
The National Highway Traffic Safety Administration is reviewing a defect petition alleging that Tesla’s Cybercab and Model Y vehicles operate above posted speed limits in active school zones. The preliminary evaluation covers approximately 1,000 model-year 2026 vehicles deployed across Tesla’s commercial robotaxi testing and service operations, and it rests on submitted video documentation that the petition says shows the automated driving system failing to adhere to dynamic school zone speed thresholds.
The agency has not yet decided whether to open a formal defect investigation. It is, however, maintaining a separate probe into the engineering documentation and technical self-certification validation Tesla used for the Cybercab, a vehicle built without pedals or a steering wheel. Tesla’s service remains operational in limited trial markets, with periodic remote or onboard safety monitor supervision.
The scrutiny echoes actions previously taken against a competitor. NHTSA and the National Transportation Safety Board opened formal inquiries into Alphabet subsidiary Waymo after fleet software anomalies involving pedestrian safety perimeters and school bus stop-arm recognition, reflecting heightened regulatory attention to how automated systems behave within the operational design domains of municipal school corridors.
Waymo, Denver Survey Finds Residents Skeptical of Safety and Accountability
Public attitudes are proving as consequential as regulatory ones. A September 2026 survey by Olson Personal Injury Lawyers, conducted after Waymo’s commercial ride-hailing launch in Denver on September 1, found persistent skepticism among Denver County residents. Sixty-one percent named safety as the primary factor in their willingness to use autonomous vehicles, and 41 percent said they were unlikely to use the driverless service. Respondents rated Waymo’s overall public safety commitment 3.8 out of 10 on average, and in-ride safety 4.0 out of 10.
Interest in robotaxis, the survey found, is concentrated in specific situations rather than everyday travel. Thirty-one percent cited getting home after drinking as a use case, 28 percent named airport trips, and another 28 percent named routes through dense parking areas, compared with 22 percent who would use a robotaxi for a standard commute. Only 8 percent of respondents reported trusting the operator, and 34 percent judged the risks to outweigh the benefits.
Respondents were also specific about what would change their minds: 57 percent pointed to access to verified safety data, and 49 percent to transparent collision reporting. On liability, 74 percent said Waymo should be responsible if one of its vehicles causes a collision, regardless of how complex the automated driving system is. Concerns about onboard telematics, location tracking, and interior cameras registered at 6.0 out of 10, and 62 percent objected to the collection and retention of video footage.
Bot Auto, Lenovo, NVIDIA, Platform Selected for Driverless Trucks
Bot Auto has selected Lenovo’s AD1 autonomous driving domain controller, built on NVIDIA’s Hyperion architecture, as the production compute platform for its Class 8 driverless trucks. Integration begins in the first quarter of 2027, supporting expansion across Texas and the Southwest toward a goal of one billion autonomous miles by 2031.
The company reports handling roughly 350 commercial loads per month, with revenue up 270 percent between May and August 2026. Bot Auto projects that its autonomous driving system, including compute and sensors, will cost $50,000 per tractor, against prevailing industry estimates of $150,000 to $400,000.
The AD1 controller meets ISO 26262 ASIL D functional safety and ISO 21434 cybersecurity standards and delivers up to 2,000 FP4 TFLOPS through dual NVIDIA DRIVE AGX Thor chips running NVIDIA Halos OS. The liquid-cooled unit includes redundant safety controllers with independent dual power rails and operates in ambient temperatures from -40°C to 85°C. Bot Auto trains and simulates its neural networks on NVIDIA HGX B200, DGX B200, and RTX PRO infrastructure, with future scaling planned on NVIDIA Rubin NVL8 hardware.
PlusAI, Texas Ventures III, Form S-4 Filed for Nasdaq Listing
Plus Automation, Inc., known as PlusAI, and Texas Ventures Acquisition III Corp filed a Form S-4 registration statement with the Securities and Exchange Commission for their proposed business combination, which would list the company on Nasdaq under the ticker PXAI. The deal sets a pre-money equity valuation of approximately $800 million and offers access to up to $300 million in gross capital, combining more than $60 million in committed financing with about $236 million held in the SPAC trust account.
Proceeds are slated to fund the planned 2027 commercial deployment of factory-built Level 4 trucks running the SuperDrive autonomous driver system, developed with TRATON Group brands, Hyundai Motor Company, and Iveco Group. The company is running freight trials with Ryder and International across Texas corridors and has generated $25 million in year-to-date revenue through its HyperFoundry physical AI development and validation platform.
TomTom, DeepRoute.ai, Digital Maps Paired With AI Driving Stack
For automakers deploying driver-assistance systems in multiple countries, the challenge is adapting to local roads and rules without starting over. TomTom and Chinese developer DeepRoute.ai have signed a memorandum of understanding to integrate TomTom’s location data and digital mapping into DeepRoute.ai’s end-to-end AI driving platform. The combined stack targets SAE Level 2+ and Level 2++ systems, drawing on high-definition map layers, road geometry, and localized regulatory data to supplement real-time sensor perception on unfamiliar road networks.
The companies say the approach addresses differing international regulations, roadway layouts, and driving behaviors without retraining perception systems from the ground up for each jurisdiction. It is being demonstrated in a vehicle on public roads in Paris during the Paris Motor Show, and is aimed at easing the expansion of Chinese automated driving stacks into markets outside their home country.
Valeo, KE:SAI, Open-Source Autonomous Driving System Planned
Valeo has entered an open-science partnership with KE:SAI, short for Kyutai ELLIS Scalable Autonomous Intelligence, to build a fully reproducible, generalist AI driving system. The partners plan to publicly release the codebase, trained model weights, and benchmark datasets, giving OEMs, Tier-1 suppliers, and researchers an alternative to proprietary software stacks.
The work pairs the valeo.ai research laboratory in Paris with the Tübingen-based KE:SAI team, building on the KITTI and NAVSIM benchmarks and on Valeo’s DrivoR and VaVAM models. The compact architecture combines real-world driving data with simulation-based reinforcement learning and learned world models to train policies for rare, safety-critical situations, and is described as a foundation for European automotive software sovereignty.
TIER IV, Arm, Autoware Deployment on Edge Compute Planned
TIER IV, the developer behind the open-source Autoware stack, has joined the Arm Total Design for Physical AI ecosystem to deliver pre-integrated configurations of Autoware for Arm-based automotive silicon, building on work with the Autoware Foundation’s Open AD Kit. The goal is to reduce integration friction for OEMs and Tier-1 suppliers designing custom systems-on-chip and centralized electronic control units.
The initiative connects the Autoware developer ecosystem to Arm’s semiconductor design partners, allowing developers to prototype and validate modular autonomous driving workloads on standardized silicon IP. By establishing pre-validated software and hardware baselines across heterogeneous compute clusters, the framework aims to streamline commercial vehicle deployment and help scale regional Level 4 transit fleets.
TIER IV, Technical University of Munich, Autoware Research Center Launched
TIER IV and the Technical University of Munich have established the TUM Autoware Research Center by TIER IV under a multi-year collaboration. The center brings together the Chair of Automotive Technology, the Professorship of Cyber Physical Systems, and the Professorship of Autonomous Vehicle Systems. Its objectives include hybrid safety architectures that pair deep-learning driving models with deterministic, rule-based supervisory guardians; benchmarking and verification frameworks spanning simulation, offline analysis, and road testing on TUM’s EDGAR testbed and CommonRoad toolkits; and standardized workflows for adapting Autoware to new operational domains.
Building on earlier urban Level 4 evaluations with TÜV-Verband, the center is intended to serve as an academic and industrial hub aligning European software-defined vehicle standards with open-source reference implementations, supporting OEM integration across German and European commercial mobility deployments.
Taisei, TIER IV, Taisei Rotec, Selected for Japan’s Autoflow Road Project
Japan’s Ministry of Land, Infrastructure, Transport and Tourism has selected Taisei Corporation, TIER IV, and Taisei Rotec Corporation to run a multi-vehicle autonomous logistics demonstration for the national Autoflow Road initiative. The eight-day evaluation, scheduled for late October through early November 2026 at the National Institute for Land and Infrastructure Management test facility, will examine platooning, bidirectional maneuvers, lane changes, merging, and precision berth stopping across straight and curved sections of a full-scale tunnel and track.
The project seeks to balance vehicle autonomy with roadside infrastructure, lowering onboard sensor costs and easing deployment in environments where GPS is unavailable. Three experimental vehicles will run TIER IV’s Autoware-based Pilot.Auto software, integrated with Taisei’s T-Localizer self-localization system and Taisei Rotec’s tunnel surface coatings, which are engineered to improve LiDAR retroreflection. The trials build on FY2025 feasibility tests of physical space constraints, wireless power transfer via T-iPower Road, and simplified roadside perception.
Nissan, Kobe Autonomous Mobility Pilot Expands With Yokohama Remote Operations
Nissan Motor Co. has detailed a fiscal year 2026 autonomous mobility pilot in Kobe City’s Nada Gogo district, scheduled for December 8 to 18. The phase expands the operational design domain and adds smartphone-based on-demand ride hailing and dispatch, using Nissan Serena development vehicles configured for SAE Level 2 driving with onboard safety drivers.
A central goal is validating Nissan’s Remote Operations Center in Yokohama, which will monitor autonomous fleets in Yokohama and Kobe at the same time. The pilot is conducted with Kobe City, Moplus Inc., and the Nada Gogo Sake Brewers Association under initiatives supported by Japan’s Digital Agency and MLIT. Nissan is targeting paid service in 2027 and full commercialization by 2030.
Volvo Cars, NVIDIA, Automated Driving Partnership Expanded
Volvo Cars and NVIDIA have announced an expanded partnership to develop automated driving architectures for upcoming production electric vehicles, extending a collaboration that began in 2017. The system pairs NVIDIA’s Hyperion sensor and compute platform with Volvo’s own safety systems to support end-to-end navigation on highways, urban streets, and in automated parking.
The computing stack includes NVIDIA’s centralized accelerated hardware, the Halos operating system, and the Alpamayo software stack. The architecture is designed to scale across automation levels, creating a migration path toward validated eyes-off and mind-off driving modes. It is set to debut commercially on a next-generation Volvo electric vehicle platform, anchoring the automaker’s long-term strategy of centralized compute and software-defined vehicles.