Multiple Companies Unveil New Smart Connected and Autonomous Driving Technologies: Aptiv Launches Vision-Only Occupant Classification System, NVIDIA Debuts Open Robotaxi Model

General Motors has introduced the Connected Headliner Module (CHM), integrating 5G, Wi-Fi, Bluetooth, UWB, GNSS, and a processor to replace traditional shark-fin antennas and TCU architectures. This innovation reduces wiring harnesses and connectors, enhances wireless communication performance, and supports software-defined vehicles, OTA updates, and advanced driver assistance features. Teletrac Navman has launched its Electronic Braking Performance Monitoring System (EBPMS), which automatically collects braking data by connecting to a vehicle’s EBS system, generates reports to meet UK DVSA regulatory requirements, and enables predictive maintenance—reducing manual inspections and vehicle downtime. Ford Motor Company has filed a patent for a steering wheel tray that can be mounted onto the steering wheel via an attachment mechanism and folded down with the wheel to create a temporary workspace or dining surface. Ford emphasized that filing this patent does not guarantee future production. Aptiv has unveiled the industry’s first vision-only Advanced Occupant Classification System, using in-cabin cameras and AI algorithms to identify occupant status—replacing conventional seat pressure sensors. The system enables intelligent airbag control and supports over 15 functions including driver monitoring and gesture recognition, while reducing BOM costs by approximately 40%. Seoul Semiconductor announced that its HV opto-semiconductor technology is now in mass production with four leading global automakers and will expand to ten vehicle models within the year. The technology employs a single-chip high-voltage driving solution, reducing component count, power consumption, and drive costs while simplifying overall vehicle E/E architecture design. General Motors has filed a patent for an airbag with variable deployment modes that adapts its inflation pattern based on occupant size, posture, and seat position. It supports multi-position and multi-directional deployment and allows dynamic adjustment of airbag shape, pressure, and inflation zones. Ottawa Infotainment has introduced the Tucker Safety Package, integrating digital alerts, distraction prevention, automatic egress, and other safety features. It also unveiled a next-generation heads-up display (HUD) concept system based on the DragonFire operating system to help automakers reduce system integration complexity. Schaeffler and Sonatus have established a global partnership to bring edge AI into motion control solutions for software-defined vehicles. By integrating Sonatus’ Collector AI and AI Director software into Schaeffler’s cross-domain control units, automakers can continuously upgrade steering, braking, and energy management functions at the edge without hardware modifications. RTA Fleet has launched Ron360, a read-only conversational AI assistant embedded within RTA Fleet360 software. Trained on real-world fleet operational data, it allows users to query maintenance, parts, and cost information via natural language. Responses include source attribution and are governed by user permissions. Diodes Incorporated has introduced the automotive-grade smart load switch DML1012ALDSQ for power sequencing in ADAS, infotainment systems, and display clusters. It operates from 0.8V to 4V input voltage, offers a maximum on-resistance of 8mΩ, supports 6A continuous output current, and integrates protections such as inrush current control, fast discharge, and under-voltage lockout. rFpro has released AV elevate IN CABIN, a simulation software package for high-fidelity testing of driver and occupant monitoring systems in virtual environments. Featuring detailed facial and skeletal rigging, an expanded in-cabin object library, and modeling capabilities for infrared cameras and radar-like sensors, it enables simulation of micro-expressions and in-cabin energy propagation before physical prototypes are built. NVIDIA has launched Alpamayo 2 Super, a 32-billion-parameter open visual-language-action (VLA) model designed for L4 Robotaxi development. It supports 360° perception, reasoning-based auto-labeling, and meta-action outputs. NVIDIA plans to release inference code and model weights in the summer. Complementary tools—including AlpaGym, OmniDreams, and Omniverse NuRec—are also being introduced to streamline the development pipeline from data collection to in-vehicle deployment. indie has launched the edge AI SoC iND881, integrating an NPU, DSP, CPU, and multi-camera ISP. Supporting diverse sensors including infrared, ToF, radar, and LiDAR, it meets ASIL-B standards and targets edge perception applications such as automotive smart cameras, humanoid robots, and autonomous mobile robots (AMRs). Samples are now available. Lockton and Nexar have jointly introduced an autonomous driving “Human Baseline” framework, leveraging 10 billion miles of real-world driving data and 60 million critical safety events. It provides a risk index and validation platform for independent safety assessments of autonomous systems, serving insurers, regulators, and AV developers. Infineon has integrated its OPTIGA Trusted Platform Module into NVIDIA’s Jetson Thor platform, delivering hardware-rooted trust and post-quantum security capabilities for robotics and embodied AI systems. The solution supports secure boot, remote attestation, and secure OTA updates, and complies with FIPS and Common Criteria certifications.