Train, bike or car? AI methods to help choose the best means of transport depending on the current situation
Mobility services offer people a variety of ways to get from A to B. Whether bus and train, bicycle or e-scooter, car-sharing or private car: the options are steadily increasing and can contribute significantly to more sustainable, cheaper and healthier mobility. The choice of the appropriate means of transport always depends on the circumstances: Which services are available, how is the weather developing, is there a lot of traffic or limited parking at the destination? Seamless mobility means being able to choose the most suitable means of transport for each leg of the journey and to switch between them quickly and conveniently.
But what is the optimal means of transportation for my situation? How can I keep track of all the options on offers? The answer is: “By using artificial intelligence!” Karlsruhe research institutions, companies and public transport operators want to jointly develop an app that bundles all available transport data and makes well-founded suggestions to citizens about the best means of transport to reach their destination.
The focus of the project “DAKIMO”, the acronym of the German project title “Daten und KI als Befähiger für nachhaltige, intermodale Mobilität” (data and AI as enablers for sustainable, intermodal mobility), is to expand the regiomove app of the Karlsruhe transport association KVV. To this end, the partners are investigating how the potential of the many data already available, for example from mobile apps, public transport operations, and traffic and weather forecasts, can be exploited to provide users with precise suggestions for means of transport that optimally suit their needs and route, to improve services, and to reduce barriers to the use of environmentally and climate-friendly means of transport. This is intended to achieve the underlying research goal of systematically analyzing the possibilities of AI-based data evaluation in the mobility use case and methodically developing procedures and architectures. Their suitability will be evaluated in a well-founded manner on the real use case in order to assess the future potential of the approach.
Situation and demand oriented recommendations
Data from public transport operators, cities and municipalities, end users, and regiomove ports (hubs where citizens can find out about and switch between different mobility options in the Karlsruhe region) are merged with environmental information on weather and traffic. Fraunhofer IOSB uses AI to process these data sets and analyze them in such a way that they lead to individually meaningful recommendations for means of transport. In addition to technical implementation, the focus is on ensuring sustainability, user-friendliness, and data protection: CO2 emissions saved by more efficient use of means of transport are to be estimated, acceptance is to be investigated in studies with test subjects, and a comprehensive data protection concept is to be developed.
The fact that external circumstances such as the weather influence our decision as to which means of transport we want to use at any given time initially seems unsurprising. “However, if this decision is to be positive in the sense of the mobility transition, i.e., if possible, in favor of environmentally and climate-friendly means of transportation such as buses, trains, or bicycles, we must provide users with optimal information as a basis for decision-making,” says Dr. Martin Kagerbauer from the Institute for Transport Studies (IfV) at Karlsruhe Institute of Technology (KIT). Until now, traffic models, which serve as a basis for traffic policy and planning, assumed that all traffic participants were equally well informed. “In practice, however, this is not the case at all,” Kagerbauer explains. For this purpose, individual factors such as age and social status are not taken into account. "For our demand model in mobiTopp, which we developed for the regiomove project, we collect such data through surveys,” says Gabriel Wilkes of IfV. MobiTopp already maps all the routes taken by everyone in Karlsruhe. Now, KIT researchers are also integrating weather and information data to simulate citizens' mobility decisions even more accurately.
Sustainable transport planning thanks to mobility database
INIT (innovation in traffic systems SE) develops integrated software and hardware solutions for the entire process chain in public transport. One of the most important tools of transport companies is the control center software (called ITCS). This allows real-time monitoring of their vehicles and, if necessary, enables the triggering of so-called dispositive measures (such as detour). At the same time, the ITCS is responsible for distributing relevant data such as real-time information to the passenger information system and passenger numbers from the counting systems to the statistics tool. Thus, the ITCS is a valuable data provider.
In the DAKIMO project, INIT will enrich the various data that converge in the control center software with public transport-external data obtained in the project and make them interpretable for external systems. The goal is to use the resulting comprehensive mobility database to gain insights into the mobility processes of the network and to enable public transport operators to adapt their services to the intermodal environment and to promote mobility behavior of users in the sense of sustainability.
INOVAPLAN GmbH researches the integration of the results into transport planning practice, especially for cities and local authorities. Findings from the use of the diverse data sources in the project will make it possible to better answer municipal questions in transportation planning and control in the future and to expand new ideas or previously limiting possibilities in transportation modeling. “Due to an increasing availability of data, their integration into municipal transport planning practice is becoming more and more important, so we are making an important contribution here in order to plan in line with the times,” says Dr. Tim Hilgert, INOVAPLAN project manager.
Regiomove app with an open, modular system
raumobil GmbH develops sustainable, digital internet solutions in the sharing and mobility sector. The app regiomove has emerged from a modular system (mappkit) for mobility applications developed by raumobil. Characteristic of the products is that they have an open modular system, which means that new interfaces and information from new service providers can be integrated at any time. This is the basis for integrating and harnessing additional mobility-relevant data sources from the DAKIMO project, which are intended to enhance the user experience of the app through personalization as well as the provision of additional information services and provide end users with access to tailored and easily accessible mobility services.
“The DAKIMO project aims to demonstrate the extent to which AI methods can support societal concerns such as the mobility revolution, while creating sustainable, demonstrable added value,” says overall project manager Dr. Thomas Usländer from Fraunhofer IOSB. To this end, DAKIMO started work in October 2021 with a budget of around 5 million euros. Over the next three years, the researchers will drive the development and implementation of corresponding solutions in the virtual data room as well as in real traffic in the Karlsruhe region.