Friedrichshafen/Detroit. ZF is continuing on its journey to becoming a complete supplier of fully automated driving technologies. The company’s products focus on advanced, integrated systems that help enable vehicles to see, think and act. To support these systems, ZF relies on a comprehensive technology portfolio of sensors ("see"), electronics and software ("think“) as well as precision mechatronic systems ("act") in drive, chassis and steering systems. In partnership with Ibeo, ZF is working on ground-breaking innovations in the fields of sensors and electronics with the goal of developing the next generation of solid-state LIDAR sensors. Car-to-x networking can considerably broaden the range of environmental sensing, and can provide additional safety-related information. In order to process large volumes of sensor data rapidly and intelligently, ZF and NVIDIA are teaming up to equip vehicles with the computing power of a super computer. This has the potential to help make deep learning and artificial intelligence possible in automotive applications.
Many vehicles today are already equipped with advanced driver assistance systems, such as emergency braking or lane keeping assist. These systems are designed to actively intervene in an emergency situation if the driver does not take appropriate action in a timely manner. To achieve fully automated driving functions, a higher degree of technology will be required.
Driver assistance systems ultimately depend upon active mechatronic systems in the driveline, chassis, steering and brakes, to be able to control the longitudinal and transverse movements of the vehicle without direct intervention from the driver. ZF’s product portfolio therefore focuses on components that are capable of integration into advanced safety and automated driving systems. “Our aim is to enable vehicles to see, think and act – and to be the one-stop shop for these systems,” explains ZF CEO Stefan Sommer. The company is getting closer to achieving this goal with its innovations in sensor technology, including developments in 360-degree panoramic vision and lightning-fast information processing. To this end, ZF is relying not only on its own power of innovation but also on its partnerships with other high-tech companies seeking to enhance the mobility space.
Sensor data quality and quantity on the rise
ZF’s current radar sensors can be used for a variety of applications, and radar can work effectively in almost all weather conditions, including poor visibility – e.g. heavy rain. The ZF AC100 and AC1000 radars are designed to meet the different needs of current driver assistance systems, and the benefits of radar include the ability to directly measure distances and relative speeds - already central in highway driver assistance systems - and even more important in the autonomous driving functions of the future. ZF is also the world leader in front object recognition cameras, which use Mobileye’s advanced image recognition algorithms and can detect vehicles, pedestrians and other road users and can trigger Automatic Emergency Braking systems to help mitigate or avoid collisions. The cameras can also monitor lane markings, and support active Lane Keeping or Lane Centering functions in addition.
Another key technology for automated driving is 360-degree panoramic vision. In level 3 and 4 automated driving systems a high degree of environmental sensor robustness and accuracy is important, and ZF’s next generation camera and radar systems will be augmented with laser-based light detection and ranging, or LIDAR systems. ZF has been partnering with Hamburg-based Ibeo since 2016 to develop solid-state LIDAR systems for introduction in automobiles. These systems will be used to measure the distance, location and relative speed of nearby objects and to confirm and validate the data from the other onboard radar and camera sensors.
Intelligent car-to-x networking from ZF can make additional information available to vehicles, enhancing the sensor data with information about objects that may not be in the direct field of view. For example, with its new X2Safe feature, ZF recently demonstrated how a cloud-based algorithm can help to increase road traffic safety. X2Safe enables communication between vehicles, smartphones and smartwatches, and is designed to give drivers, pedestrians and cyclists advanced warning of potential collisions in city traffic. This technology can form the basis for car-to-x-, x-to-car or even x-to-x applications in the future.
Super computer brings artificial intelligence to the car
Data processing needs will undoubtedly increase exponentially in self-driving vehicles. To interpret and process all this digital information from sensors, onboard vehicle systems and the cloud, the car needs to have sufficient computing power. The architecture used in automobile electronics is traditionally made up of individual electronic control units, and these can be improved and enhanced for autonomous driving.
ZF is working with NVIDIA to develop artificial intelligence (AI) systems for the transportation industry, including automated and autonomous driving systems for passenger cars, commercial trucks, and industrial applications. Unveiled at the Consumer Electronics Show in Las Vegas, the ZF ProAI for highway automated driving is ZF’s first system developed using NVIDIA AI technology. It will enable vehicles to understand their environment by using deep learning to process sensor and camera data. In addition, the companies are developing solutions for more advanced highly automated driving—in which systems control the vehicle with greater levels of autonomy.