[241022] #Huawei #Helm.ai #Fortescue

1. Huawei wants to top the flagship car market with self-driving tech

• Huawei aims to lead the Chinese flagship car market with its self-driving technologies, powered by the HarmonyOS smart automotive software.
• Yu Chengdong, Huawei’s automotive chief, believes HarmonyOS could become the top solution in the high-end car market within five years.
• Huawei’s automotive technology is already integrated into cars priced under 500,000 yuan and is now targeting vehicles in the 1 million yuan and above market.
• Huawei collaborates with brands like Aito and BAIC, controlling the driving experience and underlying technology despite working with different manufacturers.
• Huawei’s Automated Driving System (ADS) 3.0, introduced earlier this year, uses both LiDAR and camera solutions, unlike Tesla, which relies solely on cameras.
• ADS 3.0 offers enhanced parking features and AI safety checks but is not yet L3 autonomous, though it is designed with L3 standards in mind.

   

  

2. Scania and Fortescue to collaborate on autonomous road train

• Scania and Fortescue have partnered to develop and validate an autonomous road train solution for Fortescue’s mining operations in Australia.
• The solution will be tested at Fortescue’s iron ore operations in Christmas Creek, Queensland, using Scania’s R770 prime mover platform with three trailers.
• Scania’s autonomous technology will be integrated with Fortescue’s Fleet Management System (FMS) to optimize fleet usage, reduce fuel consumption, and lower carbon emissions.
• The partnership aims to improve mining efficiency and safety while promoting sustainability.
• Fortescue’s expertise in autonomous haulage and Scania’s experience in the mining industry make this a significant project in reducing the mining industry’s environmental footprint.

Scania and Fortescue have partnered to develop an autonomous road train solution for mining, focusing on improving efficiency, safety, and sustainability in Australia’s mining operations.

3. Breakthrough Volvo lidar safety system expected to be on 2025 EX90 SUV

• The 2025 Volvo EX90 electric SUV will feature a first-of-its-kind long-range lidar system, set to activate in 2025, extending driver awareness up to 930 feet.
• Luminar’s Iris lidar, mounted on the EX90’s roof, uses an infrared laser for precise detection in all weather conditions, far surpassing traditional headlights.
• The system can detect stationary objects and people at distances greater than two football fields, reducing severe accidents by up to 20%, according to Volvo’s research.
• Unlike systems used in self-driving cars, Luminar’s lidar is designed to enhance, not replace, human drivers by providing 3D data for better accuracy.
• The system’s hardware is already installed in EX90 vehicles, with software updates pending for activation; the lidar consumes minimal energy and is cost-effective at $1,000 per unit.
• Luminar’s lidar technology is expected to feature in 25 upcoming vehicle programs, marking a significant milestone in AI and safety technology in automotive.

4. Ford unveils BlueCruise 1.4: hands-free driving time doubled with new software update

5. Helm.ai Introduces VidGen-2: Generative AI for Higher Resolution and Enhanced Realism Multi-Camera Video for Autonomous Driving

• Helm.ai launched VidGen-2, a next-generation generative AI model for producing highly realistic driving video sequences, offering 2X higher resolution than its predecessor and improved realism at 30 fps.
• VidGen-2 is trained on diverse driving footage using NVIDIA H100 GPUs and Helm.ai’s unsupervised Deep Teaching™ method, generating videos at 696×696 resolution and multi-camera support with consistent views.
• The model simulates complex driving scenarios across various environments, vehicle perspectives, and weather conditions, producing human-like driving behaviors and temporally consistent object motion.
• VidGen-2 offers automakers scalability advantages by enabling rapid asset generation for autonomous driving development and validation, reducing both time and cost.
• Helm.ai CEO Vladislav Voroninski emphasized that VidGen-2 supports a wide range of automotive sensor stacks, enhancing simulation-based training for autonomous driving and robotics automation.