Ambarella’s AI chips power cabin monitoring systems
Fabless semiconductor manufacturer Ambarella, Inc, a technology leader in AI-driven computer vision and video compression, is currently on a roll: automotive supplier Continental plans to install Ambarella SoCs in numerous driver assistance systems in the future. And there is also an Amarella success story from the Chinese car industry.
In future, Continental will offer its ADAS) on the basis of the Ambarella CV3 system-on-chip (SoC) family. A decisive feature in the selection process is the AI capabilities of the SoC; other criteria were the energy efficiency and performance of the components. Ambarella’s scalable SoC portfolio, developed for ADAS applications, complements Continental’s solutions for assisted driving and further advances vehicle automation, the Tier One said.
The more sensors installed in a vehicle, the more data is collected to capture the environment. The joint solution, with centralised single-chip processing based on 5-nanometre technology, enables the next generation of vehicles to process environment sensing from multiple sensors even faster thanks to increasing performance levels. Sensor solutions using Ambarella SoCs include high-resolution cameras, radars and lidars, and ultrasonic sensors. Here, the integrated SoC enables early data fusion, where information from different sensors is overlaid simultaneously for different modalities such as emergency brake assist, parking or cruising.
“After two years of exchanging ideas with Ambarella on various use cases, Continental is now integrating its high-performance, low-latency and low-power chipset into our assisted driving solutions when processing data,” explains Frank Petznick, head of the Autonomous Mobility business unit at Continental. “This takes our ADAS systems to the next level and helps to process the growing amount of sensor data in the vehicle. In this way, we achieve more comprehensive and precise environment sensing.”
Driver assistance systems equipped with Ambarella’s SoCs can collect more data and process it faster. In this way, they increase precision and enable a wider range of functions. The high scalability of the system-on-chip family allows vehicle manufacturers to choose the necessary performance level for their system requirements on the same vehicle architecture. Despite the higher performance and accuracy, the solution consumes less power, reports Continental This also results in lower cooling requirements, enabling sustainable energy savings while reducing system costs. An added value for vehicle manufacturers to reduce their development efforts across all vehicle types.
According to Continental, Ambarella’s scalable SoC family offers significant AI performance per watt for neural network computations, with a 40-fold performance increase compared to its previous semiconductor chip family. In addition, Ambarella integrates its image signal processor technology into all SoCs. This results in robust ADAS systems with higher environmental sensing in challenging light, weather and driving conditions for both machine and human perception: humans see through the assistance system’s sensors.
Continental plans to present its ADAS systems with the CV3 SoC Ambarella to the public at CES 2023 in Las Vegas.
Ambarella has also scored an important success in China: Chinese automotive supplier INVO is supplying a cockpit monitoring system with a chip from Ambarella, which is in series production on an SUV from Chinese volume OEM GAC (Guangzhou Automobile Group). GAC’s vehicle is equipped with INVO’s Cockpit Super Sensory Interactive System. This system uses Ambarella’s CV25AQ AI Perception SoC. The chip processes the video signals from multiple cameras – one 2 megapixel (MP) driver monitoring system (DMS) camera, one front occupant monitoring system (OMS) camera and two rear OMS cameras. In addition, it processes a “life monitoring” radar inside the cabin, along with a comprehensive layout of other integrated sensing technologies. The DMS camera uses an IR sensor to sense the driver’s behavior and status in all directions and scenarios. The OMS camera uses RGB-IR sensors, which are deployed in the ceiling lights and left and right B-pillars to ensure there are no blind spots in the cabin. The combined DMS and OMS cameras can not only track the driver’s status and provide fatigue alerts, but can also track rear passengers.
Related articles:
SoC integrates high-res image and radar processing, delivers 500 eTOPS
Mobileye develops single-chip supercomputer for autonomous driving
Imagination, Ambarella team on autonomous vehicle chips
Ambarella expands ADAS ecosystem
Ambarella, Momenta collaborate on autonomous HD mapping platform
Camera chip uses AI for monitoring applications
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
