Next July, motorists criss-crossing France may come across this astonishing spectacle: on the motorway or congested expressway, a luxury limousine moving slowly in line, while the driver will be quietly immersed in reading from his newspaper… France will indeed become the second European country, after Germany, to authorize the marketing of certain cars capable of driving in level 3 autonomous mode. That is to say without the driver maintaining the hands on the wheel, as required until now by the regulations applied to vehicles equipped with a semi-automated driving function.
This magic trick, we owe it in part to the French equipment manufacturer Valeo. With its 20,000 engineers scattered around the world and its 2 billion euros invested each year in research, an effort comparable to that of Gafam, it has become one of the world leaders in lidar (“light detection and ranging”). : a radar of the third type, considered a key piece in the development of advanced autonomous driving.
Today, only two cars are approved internationally to drive in this mode of driving delegation, where the legislation is favorable to it: the latest Mercedes S-Class and the Honda Legend Hybrid EX (Japan also allows the level 3 autonomy).
“They have one thing in common, both are equipped with Scala 2, Valeo’s lidar, coupled with a complete system of sensors”, welcomes Geoffrey Bouquot, director of R&D and strategy for the company’s autonomous car. . That day, the engineer was in Detroit, the Mecca of the American automobile, where he made the rounds of meals to praise the merits of the in-house system, the aim of which is of course to improve safety on the road: 90 % of injury accidents are linked to human factors. And, since the advent of driver assistance systems (automatic lane keeping, emergency braking, etc.), we rely more and more on technology to reduce this percentage.
We often ignore it, but, on the autonomous car, France is teeming with initiatives. Research players, such as the CEA, Inria, Vedecom or the Gustave-Eiffel University, participate in large-scale projects with manufacturers (Renault, Stellantis, Valeo, etc.), engineering companies (Akka Technologies …) and start-ups to accumulate knowledge. Valeo’s system, which took ten years to develop, has benefited from these partnerships. And he has a magical side.
Placed in the front bumper, the device scans the vehicle’s environment at 360 degrees using infrared laser beams, when a conventional radar uses waves. These rays bounce off surrounding objects and return to the lidar, whose software and artificial intelligence-based machinery interprets everything it has seen. At the same time, the lidar measures the distance of objects to the nearest centimeter and their speed if they are moving. Day or night, whether it’s raining, windy or snowing, nothing escapes his lynx eye: trees, humans, vehicles, infrastructure elements…
From this amount of information, to which are added data collected by radar sensors and cameras and a constantly updated HD map, the car makes decisions. It is then the manufacturer’s software that comes into play, with the latter assuming responsibility for the maneuver. Thanks to the combination of different sources of information, he believes that security is guaranteed. “A radar is susceptible to electromagnetic interference from high voltage lines or the profusion of surrounding metal objects and a camera can be sensitive to bad weather. But if you add a third sensor like the lidar, it is not not possible that the device invents an obstacle”, supports Vincent Abadie, vice-president and expert in autonomous driving at Stellantis. The 14-brand manufacturer plans to equip its top-of-the-range models with this type of equipment from 2024.
Of course, Valeo is not the only one to develop lidars which, according to professionals, should represent a market of 50 billion euros by 2030. A crowd of start-ups are interested in it. In Europe, big players like Continental or Bosch are very active in the niche. In the United States, Waymo, the subsidiary of Google, or even Velodyne are prowessing, especially in the robot taxi market. But on the passenger car side, Valeo is still ahead. “For the moment, we are the only ones to provide technologies of this type meeting the quality standards of the automotive industry and produced in series”, recalls Geoffrey Bouquot, the expert house. And it’s not over.
Established within the framework of the Economic Commission for Europe of the United Nations (UNECE), the standard relating to the marketing of highly autonomous vehicles currently imposes certain limits. It specifies in particular that the system can only be activated in the event of dense traffic, on the motorway and at a maximum speed of 60 kilometers per hour. It is under these conditions that the driver can devote himself to something other than driving. Like surfing the Internet, checking e-mails or watching a movie on the central screen.
But the norm could quickly change. “Within two years, we plan to extend the system to 130 kilometers per hour. The service provided will go from traffic jams to motorway travel”, announces Jérôme Paschal, chief technical officer of UTAC, a specialized group in vehicle testing and homologation.
Good news for Valeo. Because the company is already ready to follow this evolution. Last November, it presented the third generation of its lidar, the production of which will begin in 2024. Unlike its predecessor, it is capable of operating at high speed. Its performance is increased tenfold, with a resolution multiplied by 12, a viewing angle by 2.5 and a range by 3 to reach 200 meters. The equipment manufacturer, which estimates that 30% of high-end vehicles sold in 2030 will be equipped with a level 3 autonomy device, has a highway ahead of it.
Dozens of young French companies are betting on the autonomous car market
Hundreds of computers are needed to manage each function of the vehicle and the proliferation of sensors will further increase their number. The result of a partnership between the CEA and Renault, the Ile-de-France start-up Alkalee has developed a centralized computer that combines all the functions. Tesla was the first manufacturer to explore this path.
The optical sensors of self-driving cars must be operational whatever the weather. ClearDrop, a start-up from Marcq-en-Barœul (59), has developed a technology using surface acoustic waves to quickly remove water, condensation, frost or snow. The system is notably used to clean cameras and lidars.
These two young shoots have specialized in inertial navigation devices, capable of taking over when the GPS is inoperative, for example in tunnels. Equipped with gyroscopes, accelerometers, magnetometers, these systems that provide high-precision position and orientation equip submarines, tankers and aircraft.
Increasingly, the autonomous car will also be electric and will be able to recharge its battery without human intervention. Gulplug, a start-up from Grenoble, offers a clever connection system. Once the vehicle is positioned above a power source placed on the ground, the socket descends from the car and automatically connects to it using magnetism technology.
In situations where the sensors for steering the vehicle are hampered (crossroads, etc.), the quality of the connectivity is crucial. YoGoKo has developed a software platform (3/4/5G, Wi-Fi) which allows communication with other cars as well as with the infrastructure to refine guidance. The autonomous shuttles of the French Navya thus use it.
Receive our latest news
Twice a week, the automotive news in one look.
We want to say thanks to the writer of this short article for this remarkable web content
Autonomous car: Valeo’s laser will revolutionize automated driving
You can view our social media profiles here and other pages related to them here.https://www.ai-magazine.com/related-pages/