Bosch to mass-produce first automotive-grade lidar system for autonomous driving

According to foreign media reports, good things come in threes, and autonomous driving is no exception. Before safe autonomous driving becomes a reality, in addition to cameras and radars, a third technology is still needed - sensors. Therefore, Bosch is preparing to mass-produce the first automotive-grade lidar system - a long-range lidar sensor. Laser ranging technology is essential for L3 to L5 autonomous driving functions defined by SAE. The new Bosch sensor covers both long and short distances and is suitable for autonomous driving scenarios on highways and urban roads. Through mass production, Bosch hopes to reduce the cost of the technology and make it suitable for the mass market.

Only by deploying all three sensors simultaneously can the highest level of safety be provided for different scenarios of autonomous driving. This has been confirmed by Bosch research, where developers investigated all application scenarios involved in autonomous driving, from assistance functions on highways to fully autonomous driving in cities. For example, if a motorcycle approaches an autonomous vehicle at high speed at an intersection, in addition to cameras and radars, lidar is also needed to ensure that the two-wheeled vehicle can be accurately detected. In this case, radar has difficulty detecting the narrow outline and plastic fairing of the motorcycle.

In addition, it is difficult for cameras to capture clear images under strong light. Therefore, radar, camera and lidar are required to complement each other perfectly to provide drivers with accurate and reliable road condition information.

LiDAR is an important component of autonomous driving

Think of lasers as a third eye: In a lidar system, the sensor emits laser pulses and captures the scattered laser light, and the system then calculates the distance based on the time it takes for the light source to return. Lidar has the characteristics of high resolution, long range, and wide viewing angle. Therefore, long-range lidar sensors can effectively identify non-metallic objects such as stones on the road, so that the vehicle has enough time to perform driving operations such as braking or steering. At the same time, the use of lidar in vehicles will place stringent requirements on lidar system components such as detectors and lasers, especially heat resistance and reliability throughout the life cycle of the vehicle.

In developing lidar, Bosch relied on its sensor and system expertise in radar and camera to ensure seamless integration of the three sensor technologies. Bosch's long-range lidar not only meets all safety requirements for autonomous driving, but will also allow automakers to efficiently integrate the technology into a variety of vehicle models in the future.

Artificial intelligence makes assistance systems safer

Bosch is leading innovation in sensor technology for driver assistance systems and autonomous driving. Over the years, the company has independently developed and produced millions of ultrasonic, radar and camera sensors. In 2019, Bosch's sales of driver assistance systems increased by 12% to around 2 billion euros. Such assistance systems pave the way for the advent of autonomous driving.

Bosch engineers have recently used artificial intelligence to optimize in-vehicle camera technology to a new level. The camera technology can identify objects and classify them into categories such as vehicles, pedestrians and cyclists, while evaluating their movement paths. In congested urban traffic, the camera can quickly and reliably identify and distinguish vehicles, pedestrians and cyclists that are partially blocked or crossing the road, and trigger vehicle alarms or emergency braking in a timely manner.

Engineers are also constantly optimizing radar technology. Bosch's new generation of radar sensors can sense the vehicle's surroundings even in bad weather or poor light conditions. The basis for achieving this performance is the sensor's long-range detection range, wider field of view, and high angular resolution.