In an era increasingly defined by urgent environmental concerns, the persistent reliance on private automobiles in nations like Germany presents a significant challenge to achieving sustainable urban environments. Despite their substantial carbon footprint, cars remain the primary mode of transportation, underscoring a critical need for innovative solutions that can make eco-friendly alternatives genuinely attractive and accessible to the broader public. Researchers are now pioneering advanced artificial intelligence to fundamentally reshape urban mobility, promising a future where seamless, convenient, and reliable travel is possible without the perpetual demand for privately owned vehicles.
The current landscape of eco-friendly transportation, encompassing buses, trains, trams, electric scooters, and shared bikes, offers a significantly reduced environmental impact compared to personal cars. However, the inherent convenience and constant availability of private vehicles have traditionally overshadowed these sustainable options. The fundamental hurdle lies in the complexity of planning multi-modal journeys, where combining different means of transportation often proves cumbersome and unreliable, deterring potential users from embracing more sustainable travel habits.
For intermodal travel to become a truly viable and appealing alternative, the experience of switching between various modes of public transportation and shared micromobility options must be as effortless and predictable as grabbing car keys. Presently, route planning applications often fall short by not adequately factoring in the real-time availability of shared resources, leaving commuters in a state of uncertainty about whether a shared bike or scooter will be waiting for them at their next transfer point. This lack of reliable information significantly diminishes the appeal of multi-modal routes, despite their environmental benefits.
Addressing this critical gap, the DAKIMO project emerges as a pivotal initiative, spearheaded by researchers at the Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB in Karlsruhe. This collaborative effort has led to the development of a sophisticated AI-based system designed to predict the availability of shared transportation options. By leveraging live data and historical patterns, this innovative artificial intelligence aims to remove the guesswork from intermodal planning, thereby empowering commuters to make informed and confident choices about their journeys.
The forecasting capability of this AI system is meticulously crafted, calculating the likelihood of finding a shared bike or electric scooter at a specific time and location within small geographical cells and short time intervals. This predictive power is derived from extensive analysis of open data sources, including public transit information and historical data regarding the positioning and usage of shared vehicles. Project partner raumobil GmbH then ingeniously integrates these AI-driven forecasts into their intermodal routing algorithms, allowing mobility apps to recommend connections that factor in the predicted availability, optimizing the entire travel chain.
The ultimate objective is to enhance existing platforms, such as the regiomove app launched by Karlsruher Verkehrsverbund (KVV), by offering users highly customized and optimized suggestions for transportation modes. These recommendations are tailored to individual needs and the chosen route, adapting dynamically to the prevailing situation. Jens Ziehn, project lead at Fraunhofer IOSB, emphasizes that this AI forecasting feature recommends the optimal means of transportation for each segment of the route, simplifying complex intermodal travel and displaying bookable vehicles, including car-sharing options, at both the start and end of trips.
The success of the AI fusion server, which compiles and processes all data to determine transportation availability and compute intermodal routes, underscores the project’s robust technical foundation. Furthermore, the integration of this AI forecasting feature into a test version of the Karlsruhe-based regiomove app has already demonstrated its practical utility, combining a broad spectrum of transportation options for the Middle Upper Rhine Region. The enthusiastic public response, with nearly 90% of over 1,500 study participants finding the AI-based predictions helpful, further validates the system’s potential to drive positive change in urban mobility.
Looking ahead, plans are in motion to expand the forecasting model beyond its initial test phase in Karlsruhe to the entire state of Baden-Württemberg, signaling a broader commitment to integrating AI technology into public transportation infrastructure. Reinhard Herzog, who leads the Modeling and Networked Systems group at Fraunhofer IOSB, highlights the importance of incorporating these forecast probabilities into the GBFS standard for sharing vehicles, which has already been accepted by MobilityData. This standardization is crucial for ensuring the widespread adoption and interoperability of this cutting-edge solution across various routing applications.
These research findings collectively confirm the profound impact AI-based methods can have on supporting the mobility transition and significantly contributing to global climate action. By making eco-friendly travel more reliable, flexible, and convenient through intelligent route planning and shared mobility solutions, the DAKIMO project paves the way for a future where urban dwellers can confidently choose sustainable alternatives, thus fostering a greener, less car-dependent society.
