Researchers have developed a new approach to automate the tiresome search for a parking space by combining the techniques of autonomous driving and the Internet of Things. The concept, developed by the Institute for Transport Systems Technology at the German Aerospace Center (DLR) as part of the EU's Autopilot project (Automated Driving Progresses by the Internet Of Things), allows the driver to park his robot car at a specified point, such as the office and send it via app for parking.
The IT system behind the sub-project "Automated Valet Parking" knows where there are free parking spaces nearby, navigates the vehicle there and parks it – backwards smoothly if necessary. The information about free parking spaces comes from a networked and fully autonomously flying drone as a component from the Internet of Things (IoT). It flies a set course at a height of two or three meters and collects relevant data with its four cameras. With these, the system can make statements about where the cars entrusted to it can be parked, determine the optimal route and also discover obstacles along the way.
Successful practical test
The used one Ardea drone the DLR Institute of Robotics and Mechatronics originally manufactured to explore the planet and disaster areas. In this context, it should help, for example, to track down people in buildings that are not safe to enter or that have already partially collapsed. Therefore, it flies relatively low and independent of satellite navigation data.
The IT platform in the background enables all integrated devices and elements to be addressed via appropriate interfaces. The open source solution, which is independent of individual providers, also manages the data traffic. In a practical test at the DLR site in Braunschweig, the scientists successfully demonstrated that all of the components work together virtually and in real life in a quarter that is, however, relatively busy. For example, Bosch and Daimler had already started parking "as if by magic" in a Stuttgart multi-storey car park.
As part of from autopilot have over 40 partners from research and industry such as Continental, Gemalto, T-Systems, TomTom, Huawei, the French PSA Group and IBM for around three years in new, IoT-driven autonomous driving functions and services as well as the required IT architectures worked. In addition to drones, they also integrated traffic lights and other sensors into solutions for "smart mobility".
Scientists involved also used the initiative, which was funded by the EU with almost 20 million euros, to investigate what requirements, expectations and concerns users of autonomous technology have. To this end, they conducted an online survey, accompanied practical tests on site and explored the experiences of the pilot users. "The response was generally positive," summarizes DLR researcher Viktoriya Kolarova. "It is important that the offers are easy to book or use." The more information there is on how the technology works, the greater the trust.
According to the expert, the participants found the IT security, reliability and data protection issues to be particularly problematic. For autonomous parking, the respondents also wanted an option to be able to find out where their vehicle is at any time.
. (tagsToTranslate) Ardea (t) AutoPilot (t) Drones (t) EU (t) Parking (t) Parking