The experts agree: The mobility of the future will be shaped by Artificial Intelligence (AI) and digitization. Autonomous driving will make transportation more efficient, safer, more environmentally friendly and also cheaper. It will open up a whole new world of possibilities in mobility and logistics. Targeted testing in real-life test environments and demonstrating this technology to the public are both important prerequisites for this development to find acceptance.
The newly-approved research project “BeIntelli” aims to develop the possibilities of AI for tomorrow's mobility based on platform economy and test them under real-life conditions. A project showcase will furthermore allow the public to experience AI applications in mobility first-hand. “BeIntelli” is an interdisciplinary collaboration of researchers and partners from industry. The project will also lay the foundation for the establishment of a Berlin center for tangible AI and digitization in mobility research.
The consortium is led by Professor Dr. Sahin Albayrak, head of the Distributed Artificial Intelligence Laboratory and the Chair of Agent Technologies in Business Applications and Telecommunications at TU Berlin. Funding for the project amounts to around 17 million euros, with approximately 13 million provided by the German Federal Ministry of Transport (BMVI) and the remaining 4 million contributed by the project partners. The new research builds on the preceding project DIGINET-PS, in which TU Berlin and its project partners set up a digital urban testbed for automated and connected driving on Straße des 17. Juni. “BeIntelli has five core goals,” explains Sahin Albayrak: “One is to create technological innovations for everyday life. We aim to research, develop and test novel ideas and AI-based approaches as part of a holistic approach to mobility, including autonomous and connected driving, autonomous public transportation and automated last-mile logistics solutions. We are therefore developing a scalable software stack at TU Berlin – the AI Mobility Operating System – based on Artificial Intelligence, machine learning and distributed intelligence.”
The second core goal and prerequisite for the utilization of this software is to provide the necessary infrastructure. The existing digital testbed along Straße des 17. Juni will thus be extended and expanded in the direction of Kurfürstendamm and the Reichstag building. This includes the installation of the corresponding sensor technology along the roads, the provision of an edge infrastructure and the establishment of a 5G communication infrastructure.
The third goal is to test and validate autonomous vehicles on the testbed. In this new project, it is not just the roads that are being digitized. The vehicles that will drive on this testbed will also be enabled with functions for autonomous driving. Three test vehicles will be equipped with the necessary sensors and cameras in order to register their entire surroundings. Furthermore, the software developed at TU Berlin will operate the control technology of the vehicles, enabling true autonomous driving. In addition to passenger cars, the researchers will also equip a delivery vehicle and an “explanation bus” with this technology. “With these vehicles we can then run tests under everyday operational circumstances and get to know the real-world requirements,” says Dr. Jan Keiser, member of Sahin Albayrak's team and sub-project manager.
The fourth step is to establish a platform economy for the new mobility. “The data from the smart infrastructure and vehicles as well as the AI models and services will be processed and made available in order to promote the establishment of new business models and ecosystems. We will host corresponding events and competitions to support this,” explains Marc Augusto, sub-project manager for the development of AI applications.
A central focus of the project is to involve the public. An important core goal is therefore the public presentation of autonomous mobility in a real-life environment. “We want to not only inform the public but also encourage them to try out this new form of mobility so that the benefits and background of AI-based mobility become apparent to all road users right there on the test route,” explains Augusto. This is where the so-called explanation bus comes into play, a mobile real-world laboratory. The interior design will include large displays that visualize for the passengers what the sensors along the road and on the bus are doing as well as how the software operates the control technology. The bus can be used for public events. “However, we also plan at certain times to use the bus as a regular means of transportation on the test route Passersby are then invited to use it free of charge. Trained personnel on the bus will be available for explanations and discussions with the public,” explains Dr. Axel Heßler, sub-project manager for the development of the AI platform.
The demonstration of autonomous driving will require special permits for the test vehicles. Trained security personnel will continue to supervise the ride and be able to take control of the vehicle at any time.
Researchers from TU Berlin joining Sahin Albayrak's academic chair in the project are Professor Dr. Frank Straube, Chair of Logistics; Professor Dr. Søren Salomo, Chair of Technology and Innovation Management; and Professor Dr. Friedel Gerfers, Chair of Mixed Signal Circuit Design.
The following external partners are also involved in the BeIntelli project: GT-ARC, VMZ Berlin, ADAC BBR, IAV GmbH, Bezirksamt (district authority) Charlottenburg-Wilmersdorf / Berlin, Cheil Germany GmbH, DB Regio Bus Ost GmbH, Continental Automotive GmbH, TÜV NORD Mobilität GmbH & Co. KG, Berliner Verkehrsbetriebe, T-Systems International GmbH.