Page 14 - EE Times Europe Magazine – June 2024

14 EE|Times EUROPE

AUTONOMOUS VEHICLES | SENSORS
Automated Driving Is Transforming the Sensor

and Computing Market

By Pierrick Boulay, Yole Group

hile most OEMs are focusing on hands-off of regulation and the resulting liability issues prompted OEMs to stay
driving applications, only a handful are with hands-off applications. Of course, hands-off driving has become
quite significant, and some OEMs are more advanced than others. Tesla
implementing eyes-off driving features. has led for many years, though some others are now catching up. How-
WBefore going deeper into the eyes-off ever, the main competitors in this domain are not only global European
or U.S. OEMs but also Chinese automotive players.
scenario, this article examines the technology and current
Chinese OEMs, especially new ones like XPeng, Li Auto and Xiaomi,
market status. are pushing very hard to implement new ADAS features. All are focus-
ing on the Navigate on Autopilot (NOA) feature. Their approach is
Eyes-off driving has been around for years but appears much more based on a similar set of sensors, including cameras, radar and LiDAR
complex than OEMs anticipated. One challenge is the high number of used by global OEMs, but those players are going a step further.
sensors required not just around the car but also inside it, to monitor A detailed analysis reveals they are embedding more than 20 sensors,
the driver’s attention. Apart from sensors, computing requirements are many more than necessary to enable hands-off driving features.
increasing so that all the data from the various sensors can be gathered Computing is also part of the game and follows the same evolution.
and fused to deliver an accurate view of the car’s surroundings. That Computing power, usually supplied by Nvidia, ranges from
alone requires a considerable investment from the entire supply chain, 500 TOPS to 1,000 TOPS, while the Mobileye Eye Q5 performs at
not only in terms of money but also in terms of resources, knowledge, about 25 TOPS. Some argue that Nvidia is using brute force to process
time and regulation. Recent difficulties encountered by robotaxi com- all the data generated by the sensors, and this solution is not as opti-
panies have not helped gain customers’ trust in autonomous driving. mized as Mobileye’s SoC. In any case, Nvidia’s platform is much more
Another concern for OEMs is the extent of their responsibility if an open for OEMs, while Mobileye’s is still considered a black box, even if
accident occurs while a vehicle is in automated driving mode. In 2017, it is improving.
Audi integrated a LiDAR, supplied by Valeo, in its Audi A8 sedan, but in Facing this market evolution, Chinese OEMs are putting a lot of
2020, the German carmaker gave up Level 3, eyes-off automated driv- effort into these new developments, waiting for new regulations that
ing developments because of the lack of a legal framework. This lack could come shortly, as testing permits for automated driving are

JUNE 2024 | www.eetimes.eu