Tensor Open-Sources OpenTau
Tensor announced the open-source release of OpenTau (τ), a scalable and reproducible AI training toolchain designed to advance Vision-Language-Action foundation models, a core technology for next-generation Physical and Embodied AI systems. OpenTau brings industrial-grade VLA training capabilities—such as heterogeneous dataset co-training, discrete action modeling, specialized reinforcement learning pipelines, and techniques to improve convergence and reduce overfitting—to the broader research community, breaking them out of closed, proprietary environments. By open-sourcing OpenTau, Tensor aims to promote transparency, collaboration, and faster innovation across applications including autonomous driving, robotic manipulation, and navigation, inviting researchers and developers worldwide to build, experiment, and contribute to the future of Physical AI.
Karsan Next-Generation Autonomous Transit
At CES 2026, Karsan showcased its vision for intelligent and autonomous public transportation with the global debut of Karsan AI (Autonomous Intelligence) and its next-generation autonomous vehicle lineup. Emphasizing that the future of mobility is not only electric but intelligent, Karsan highlighted its end-to-end autonomous solutions, strong global partnerships, and real-world deployment experience across North America, Europe, and Asia. A key highlight was the Autonomous e-JEST, newly introduced to the North American market, which builds on years of field experience and has already secured its first regional orders. With more than 100,000 miles of autonomous driving, over 60,000 passengers transported, and a 99.5% success rate, Karsan demonstrated that autonomous public transport is no longer experimental but a scalable, real-world solution shaping the future of urban mobility.
Visteon High-Performance Compute for AI-Defined Cockpits
Visteon showcased the production-ready specifications and OEM deployments of its High-Performance Compute solution built on the Snapdragon Cockpit Elite platform, highlighting a major step toward centralized, software-defined vehicle architectures. Developed in collaboration with Qualcomm Technologies, the platform delivers significant gains in CPU, GPU, and AI performance, enabling edge-based generative AI and highly personalized, responsive in-vehicle experiences. Featuring Visteon’s cognitoAI Concierge digital assistant, advanced multi-user capabilities, and a scalable architecture that consolidates multiple ECUs, the solution allows automakers to reduce system complexity and costs while delivering premium, AI-defined cockpit experiences across multiple vehicle lines.
Cinemo Agentic AI
Cinemo announced the Cinemo ICO™ Agent Store, an embedded and cloud-native agentic AI platform that enables OEMs to instantly deploy, manage, and update intelligent agents across vehicles and fleets. Designed as a unified Agent-as-a-Platform, the store offers a curated range of automotive-grade AI agents and tools—such as productivity, trip planning, reservations, food ordering, and smart home control—powered by Cinemo and third-party services. By acting as a one-stop shop for agentic AI, the platform removes the need for individual integrations, contracts, or app downloads, giving OEMs real-time control and flexibility to deliver the best in-car user experiences.
Oshkosh AI-Powered Collision Avoidance System Wins Award
Oshkosh Corporation was named a winner of the CES Picks Awards for its Collision Avoidance Mitigation System (CAMS), recognized in the TWICE category for innovation showcased at CES. CAMS is the first purpose-built, AI-powered collision anticipation system designed to protect firefighters and roadside responders, using radar, computer vision, and AI to detect oncoming traffic and issue real-time alerts. After extensive field testing with fire departments, Oshkosh is scaling the technology to support EMS crews, police officers, and tow operators, with future plans for mobile highway-deployable units. Developed by Pratt Miller Motorsports and tested on Pierce fire apparatus, CAMS adds race-proven intelligence to emergency response, reinforcing Oshkosh’s continued leadership in safety and innovation for everyday heroes.
Motive AI Dashcam Plus
Summary: Motive introduced AI Dashcam Plus, a first-of-its-kind, all-in-one dash cam powered by the Qualcomm Dragonwing™ QCS6490 processor, delivering three times more AI compute and the ability to run over 30 AI models simultaneously at the edge. Featuring dual forward-facing stereo vision, advanced sensor fusion, automated license plate recognition, and hands-free two-way communication, the unified device is designed to detect risks faster, reduce false alerts, and prevent more collisions. By combining dash cam, vehicle gateway, voice assistant, and enterprise-grade connectivity into a single platform, AI Dashcam Plus cuts installation time, improves reliability, and supports a shift from reactive to proactive, AI-driven fleet safety.
Agnik MLP Vehicle Analytics & Agentic Control
Summary: Agnik announced the expansion of its vehicle analytics portfolio with deep machine learning–based solutions designed to power connected vehicle products and real-time agentic control systems. Built on decades of research in distributed machine learning, Agnik’s approach combines transformer-based models with proprietary LLM architectures that leverage local, distributed algorithms to analyze large-scale, multi-modal vehicle data efficiently. Developed by Agnik Sparks Lab, the scalable platform supports unsupervised and supervised learning, fine-tuning, reinforcement learning, and real-time inference while reducing power consumption, enabling deeper insights into the vehicle ownership lifecycle and new AI-driven capabilities across automotive and IoT applications.
Geely Unveils Full-Domain AI 2.0 and G-ASD
Geely Auto Group showcased two major innovations: Full-Domain AI 2.0, a unified vehicle-wide AI architecture that integrates computing, data, and models across cockpit, chassis, safety, and driving systems, and G-ASD (Geely Afari Smart Driving), its next-generation intelligent driving system designed for high-level autonomous driving. These breakthroughs aim to create a “super AI brain” for vehicles, enabling real-time collaboration across domains, enhanced safety, and confident driving in complex scenarios. Geely positions these technologies as foundational steps toward “Super Intelligence” vehicles by 2030, combining humanlike awareness, proactive service, and continuous evolution.