01. SEPTEMBER 2020 Tackling Empty Runs With Data Science: "KIMoNo" Research Project Studies Non-urban Mobility
Passau, September 01, 2020—Cars with only one passenger, half-full public buses, and unused truck loading space: These examples, drawn from German freight transport statistics*, vividly demonstrate how inefficient passenger transport and, above all, logistics are in Germany. The country recorded around 155 million empty runs over around 6.5 billion kilometers. Currently, there is a lack of solutions for rural areas in particular. This is precisely where the KIMoNo research project (AI-based, cross-type mobility optimization in non-urban regions) comes in, coupled with the technology and expertise of data specialist One Data from Passau. A major challenge of KIMoNo is the structuring of diverse and heterogeneous data.
While there are many studies on improving mobility in metropolitan areas, little attention has been paid to the requirements of rural regions. To change this, One Data is participating in a two-year joint project funded by the Federal Ministry of Transport for one million euros. Various chairs of the University of Passau and other partners** are involved in the implementation. After Federal Transport Minister, Andreas Scheuer, announced the funding decision on June 17, the first project phase began at One Data. Within the scope of the project, One Data is investigating two key areas: “Decentralized AI infrastructure for local mobility applications” and “AI-based methods for optimizing mobility movements”.
The district of Passau/Bayerischer Wald was chosen for the pilot study, as KIMoNo aimed to determine how mobility in this rural area could be improved with artificial intelligence (AI). Based on these results, a prototype optimization platform would be developed as a digital tool, empowering rural areas to catch up in terms of mobility—while consuming as few financial resources as possible.
Tracking Mobility Patterns with Artificial Intelligence
With its One Data data platform, One Data provides the central data processing and analysis platform for the diverse data that is collected, processed, and analyzed within the scope of KIMoNo. Dr. Emanuel Berndl, Product Community Manager and internal KIMoNo Project Manager at One Data, considers data preparation to be an important starting point for project success: “Experience has shown that this step demands a major part of the overall effort. The data comes from different sources, has different formats and a varying quality of information. All this needs to be standardized using appropriate data science expertise, and without leaving any remaining data holes.”
This data diversity is a result of the holistic nature of the project. To gain a comprehensive view of non-urban mobility, KIMoNo follows a cross-type approach: Here, multiple forms of mobility are taken into account, from individual cyclists and autonomous cars to networked truck fleets. One aspect is analyzing data to identify empty runs or redundant trips in passenger and freight transport. Experience with technologies, such as computer tomography, is vital in identifying loading space utilization. Another approach addresses the cross-sector networked optimization of routes based on road conditions. For this purpose, edge computing on smartphones processes user data. It may recommend a route that is longer but less damaging to the vehicle, thereby reducing maintenance costs.
Predicting Traffic Flows Accurately for the Long-Term
In addition, One Data will help optimize freight logistics as well as passenger transport by using appropriate forecasting: “Whether supermarket deliveries or the number of cabs required, with One Data we can more precisely calculate demand weeks in advance based on historical data enriched with events such as public holidays or weather data. We had already gained extensive experience during a corresponding project, in which we calculated a cross-industry forecast of goods volumes half a year in advance”, says Dr. Emanuel Berndl about the project-specific expertise of One Data.
Especially in rural areas and their connection to trans-regional transport networks, there is a high potential for optimization by using an appropriate tool: “Avoiding unnecessary transport and efficiently controlling mobility and logistics offers wide-ranging benefits for society, the environment, and the economy. In addition to financial savings, the use of data science and AI via One Data can contribute to safe, resource-saving, and thus sustainable transport concepts,” says Andreas Böhm, Founder and Managing Director of One Data, explaining the relevance of the project.
The project will result in a so-called Blueprint, a web-based prototype based on the One Data platform that can be used for similar use cases in the field of mobility. Making mobility more sustainable, safe, and efficient through digitalization is on the agenda at an informal meeting of EU transport ministers to be held in Passau in October 2020. The project partners will present a comprehensive project report at the end of the two-year project period.
* Source: BMVI https://www.bmvi.de/SharedDocs/DE/Artikel/G/amtliche-gueterkraftverkehrsstatistik.html
** Among the KIMoNo project partners are the University of Passau (Chair of Data Science headed by Prof. Dr. Michael Granitzer, Institute FORWISS headed by Prof. Dr. Tomas Sauer, Centrum für Marktforschung (CenTouris) headed by Dr. Stefan Mang, Department of Communication and Marketing (Dept. KM) headed by Anja Schuster and the Faculty of Philosophy headed by Prof. Dr. Malte Rehbein), Fraunhofer EZRT (Development Center X-Ray Technology, a division of the Institute for Integrated Circuits (IIS) and Cartesy (part of the Micro-Epsilon Messtechnik GmbH & Co. KG, Ortenburg)Download press release
About One Data
One Data is a leading company for the use of artificial intelligence (AI), founded in 2013. It enables its customers to monetize data quickly and sustainably and to digitize business models. As an agile provider of AI-based data products, One Data uses the latest technologies and algorithms from the fields of machine learning, deep learning and AI.