Automated Driving Chauffeured safely while you relax

The car drives, the driver relaxes – this was long considered something out of science fiction, but thanks to the newest electronic assistance systems in the car, it is close to becoming a reality. A look forward to year 2025 will reveal more.

Whether in inner-city, stop-and-go traffic, or stuck right in the middle of a dense convoy of cars filled with families heading off on vacation, or rushing off to a business meeting on short notice without much preparation time, almost every driver has probably wished at least once for an “auto pilot.” Now, this wish might just be coming true – gradually with the current and coming generation of electronic assistants. These Driver assistance systems such as ABS, ESP, and traction control have been around for a while. In the past few years, far more advanced applications have also emerged.

These Advanced Driver Assistance Systems (ADAS) do not just watch the driver and the vehicle, rather, equipped with cameras and radar, they also take into account the surroundings – for example, other road users, signs, highway markings – in order to issue hazard warnings and, if necessary, take action. A decisive factor for ADAS and their further development is the close, intelligent networking of systems, both with one another as well as in the form of vehicle-to-X communication. The “X” represents a placeholder for everything that cars of all types will permanently replace in the future – for example, for other road users, infrastructure, such as traffic light or traffic guidance systems, navigation data, telematic services or the Internet. These advances are ultimately making automated driving possible.

Seeing, thinking, acting

Sensor technology will be a major player in the future because it will have to be able to record the car's surroundings and traffic activities in a more detailed and broad-ranging manner. To do so, ADAS require a kind of set of eyes, or to be more precise, infrared and image-based systems. Among the first are the comparatively inexpensive infrared laser sensors (Light Detection And Ranging – LiDAR), most of which are installed behind the windshield and measure the distances by sending out light waves and then receiving them again as reflections of objects. However, these sensors are only suitable for low speed ranges – they are thus optimal for city emergency brake assistants. Should greater distances also be covered, then sensors with radar (Radio Detection and Ranging) are the best choice – and are housed in the front area of the vehicle. These also make adaptive cruise control and traffic assistants for stop & go traffic possible. The third option for infrared-based systems is probably the most well-known and widely used: ultrasonic sensors for Park Distance control.

As soon as the electronic system identifies what kind of object is in the vehicle's surroundings, the image-based systems come into play. These can be either a digital video camera or an infrared or thermal imaging camera for highly precise night vision functions. Image processing allows the objects in the surroundings to be interpreted: Already the “one-eyed” mono cameras are making lane departure and tailgate warning systems, lane tracking assistant, and road sign recognition possible. Stereo cameras – generally consisting of two mono solutions – are the key to having cars see three-dimensionally – which helps them very precisely determine the directions of movement of pedestrians, for example, and to calculate them in advance. Particularly important in the process is for the infrared and image-based systems not to mutually exclude one another, rather complement one another in their respective strengths and functions – by means of a sensor or data fusion.

The enormous volume of the most diverse sensor information merges together into the electronic control units, which, together, form the brain of the car, so to speak. For its commands to be able to be executed, actuators are located in different areas in the car. They convert the electric ECU signals into mechanical movements, actuating, for example, the brakes on each wheel exactly as requested by the ESP or emergency brake assist, or they turn the wheels in just as the emergency steering assist calculated for safely steering around the obstacle.

Fewer accidents, more time

There is one primary reason why Advanced Driver Assistance Systems should assume more and more tasks as quickly as possible: They are considered the only option for achieving “Vision Zero” or zero accidents. The benefits of automated systems are therefore obvious. These systems do their jobs always 100-percent reliably and with complete focus, there are no issues of exhaustion, distraction, or health. Besides this enormous safety benefit, there are also the already-mentioned comfort and time gains. The driver can relax, do some work, or even surf the Web safely, for example, while on the road in automated mode.

Highly automated driver-assist systems will make driving more comfortable and safer, especially on highways and freeways.

Moreover, automated means coordinated and efficient. With heave traffic volumes or obstacles along the route, cars could warn each other like commercial vehicles do. They could also then mutually agree on the optimal way to proceed to prevent hazardous situations or traffic jams or to mitigate their impacts. In addition, because transport vehicles constantly communicate with infrastructure – and vice versa – they also know how to perfectly leverage traffic light changes: The cars select their speed independently so that they can hit a wave of green lights as they drive through the city. This saves additional time, reduces vehicle fuel and power consumption, especially, and additionally reduces wear and tear.

In the last foreseeable development stage, fully automated driving will evolve into autonomous driving. Then no one at all will have to sit behind the redundant steering wheel – if there is one at all. But back to the present day, where mainly legal questions are standing in the way of autonomous driving, for example, who is liable if the car is going too fast by itself, because it did not recognize the speed limit? To whom does the collected data belong, how are the data protected, and who can access the data? Will it be possible for cars to be "driven" starting from a certain development stage in the assistant systems and without the occupant having a driver's license? And on what basis does the system decide, in the face of unavoidable accidents, which is the lesser of many evils? Not until these and numerous other legal, political, and ethical details are clarified will cars be allowed to do all that they are technologically already capable of thanks to ZF.

Pictures: Detlef Majer; Graphics: ZF, Sascha Bierl, Fraunhofer Institut

Further related articles