[241122] #Waymo #TRATON #RideFlex

1. Waymo explores using Google’s Gemini to train its robotaxis

• Waymo is advancing its autonomous driving technology by integrating Google’s multimodal large language model (MLLM) Gemini into a new training model called EMMA (End-to-End Multimodal Model for Autonomous Driving).
• EMMA processes sensor data to generate future trajectories, enabling Waymo’s driverless cars to make decisions about navigation and obstacle avoidance in complex environments.
• Traditional autonomous driving models have separate modules for perception, mapping, and planning, but Waymo believes that MLLMs can help overcome module communication issues and better handle novel environments.
• MLLMs like Gemini provide a broad “world knowledge” from internet data and exhibit “chain-of-thought reasoning,” which improves decision-making in unexpected scenarios.
• Although promising, EMMA has limitations, such as difficulties processing 3D sensor inputs and handling large image frames, which require further research before being applied on public roads.
• Waymo is cautiously optimistic, acknowledging that more research is essential to resolve current limitations and ensure safe, reliable large-scale deployment of MLLM-powered autonomous vehicles.

2. Autonomous Trucks: TRATON Announces Global Release Of Plus SuperDrive

• Plus and TRATON GROUP brands Scania, MAN, and International have completed the first phase of collaboration to deploy autonomous trucks, reaching a milestone with the Beta 5.0 release of Plus’s Level 4 SuperDrive™ software.
• Public road testing of SuperDrive™ was conducted across Europe and the U.S., with mapped hub-to-hub routes in Texas for phased commercial rollout.
• The deployment was coordinated by teams in Sweden, Germany, and the U.S., emphasizing global collaboration for software testing, integration, and reliability across continents.
• Texas will serve as a strategic base for launching hub-to-hub autonomous trucking operations in the U.S., with expansion planned to other states.
• The collaboration highlighted four key processes: defined testing, feature metrics for autonomy readiness, global alignment, and consistent feature performance across regio.

Plus and TRATON have achieved a major milestone with the debut of SuperDrive™ software, starting their autonomous truck testing in Europe and Texas.

3.Innoscience expands its 100V automotive-grade portfolio for the automotive LiDAR market

• Innoscience Technology expanded its portfolio with two 100V automotive-grade GaN-on-Si devices, INN100W135A-Q and INN100W800A-Q, for applications in LiDAR, high power density DC-DC converters, and Class D audio.
• The new devices are AEC-Q101 certified and optimized for L2+/L3 assisted driving systems, with switching speeds up to 13 times faster and improved recognition distances of 200/300m, essential for ADAS and autonomous driving.
• The compact WLCSP packaging of the devices enhances power efficiency and reduces power loss and temperature rise, outperforming traditional silicon solutions in critical automotive applications.
• Both GaN products are now in mass production with batch orders being fulfilled; detailed specs and samples are available on Innoscience’s website.

4. Aurora Innovation delays commercial autonomous truck launch to 2025

5. RideFlex, an autonomous software startup, announced on the 31st that it has secured a 26 billion investment

• Autonomous software startup RideFlux secured 26 billion won in Series B investment, bringing its total funding to 55.2 billion won.
• Major investors include Atnum Investment, Uanta Investment, Korea Investment Partners, and others.
• RideFlux specializes in full-stack autonomous driving software and received Korea’s first temporary driving approval for unmanned autonomous vehicles in June.
• The company operates autonomous services, including “Nemoride” in Jeju and the world’s longest autonomous bus route between Jeju City and Seogwipo City.
• RideFlux is developing a self-driving freight truck for a 61.3km route in partnership with JIAT, with plans to expand services to cities like Seoul and Busan.
• The investment will support talent acquisition, software development, and expansion into the self-driving logistics sector.