Hexagon Introduces cloud based ADAS Software Testing
Hexagon introduced its VTDx cloud based system to train, test and validate ADAS software. Physics based simulation is used to model the environment and thousands of test cases can be generated and run automatically against the “software under test”. Continuous integration and testing can significantly speed up releases and improve the quality of the release, which will be especially critical for SDVs.
LeddarTech and Texas Instruments collaborate on cost effective ADAS solutions. First OEM Design Win for LeddarVision ADAS Solution
Using the TI TDA4V processor, LeddarTech is offering up to a 40% cost reduction and 2x performance increase over existing systems. Early fusion is the key to improved performance on object detect and recognition as well as optimized algorithms for the TI processor.
LeddarTech Announces First OEM Design Win for LeddarVision ADAS Solution:
The customer is one of the world’s leading commercial vehicle OEMs (original equipment manufacturers) has selected LeddarTech as the fusion and perception software supplier for their advanced driver assistance system (ADAS) program for 2028 model year vehicles.
BMW boosts ADAS capability with help from Qualcomm
Increased processing power from the Qualcomm Snapdragon Ride processors allow BMW to offer a better understanding of road conditions and smoother lane changes on its upcoming Neue Klasse models. AWS was used to help develop the software, collecting data from thousands of kilometers of driving and validating the quality of the data and allowing multiple partners to collaoborate on the development in the cloud.
Qualcomm and Hyundai Mobis Power Next Generation ADAS and Digital Cockpit solutions
This is another example of a single SoC being used to power both the IVI/Cockpit and ADAS systems in a vehicle, using Mobis’s software and systems capability to deliver enhanced performance, safety and efficiency.
Visinex closes Y100M funding round
Horizon Robotics has been quite successful in China at selling AI processors into the L2/L2+ ADAS Market. There are a number of other start-ups targeting this market, including Visinex that is developing Vision+AI SoCs. The base SoC offers 24 TOPS performance and supports up to 7 cameras and driving/parking algorithms. An enhanced version offers 40 TOPS performance and supports “Commute” NOA. The first systems using this processor are expected to go to production at the end of 2025.
Smart and DeepRoute.ai Partnership
Smart (a Mercedes/Geely JV) and DeepRoute.ai are co-developing autonomous driving solutions, including handling complex cases such as roundabouts, sharp turns, narrow congested lanes and unmarked lanes. DeepRoute has raised more than $500M of funding since 2019 and has also partnered with other major Chinese automakers including DongFeng. Its software has focused on power efficiency and is also targeting the European market.
Micron and Black Sesame launch new ADAS Solutions
Black Sesame, based in Wuhan, develops AI chips and was funded by SAIC, Tencent, Bosch, Dongfeng, Nio, Geely and others. Customers include JAC, FAW Dongfeng, Baidu and JICA. Their latest A2000 chip is developed for Urban Autopilot and interfaces to Micron’s automotive grade LPDDR5 memory. The China SoC developers are continuing to gain share in their local market for ADAS solutions and Black Sesame is already one of the larger players here. For the first 6 months of 2024 they did 180M RMB in revenue (Approx $25M).
Horizon Robotics and iMotion form Strategic Partnership
Horizon’s Journey 6 SoC is already designed into 100 vehicle models for ADAS functions and is projected to ship over 1 million units by the end of 2025. iMotion develops front camera systems and “intelligent driving” domain controllers as well as the software for ADAS.
