Artificial Intelligence in the Ingolstadt Transportation System (KIVI)

Long-lasting traffic jams on the roads, dangerous traffic situations, frequent accidents – every day in every city there are more than enough critical situations in urban traffic. Artificial intelligence (AI) will provide a remedy here in the future. As part of the project “Artificial Intelligence in the Ingolstadt Traffic System” (KIVI), funded by the German Federal Ministry of Digital Affairs and Transport (BMVI), research is being conducted into the use of artificial intelligence (AI) to improve the traffic system and road safety in the city of Ingolstadt. The project has been running since October 28, 2020, and ends on October 27, 2023, after three years of operation, and represents an important step toward more efficient and safer transportation design.

Nikolai Zotov

Science Editor

Nikolai Zotov

Science Editor

Challenges in the current transport system

Despite progressive digitization and networking of transport systems, a lot of information is still missing to create a coherent overall picture. While individual movement data of single road users is available through navigation devices, cell phone data and GPS tracking, a comprehensive picture of the entire traffic situation is missing. In particular, there is often insufficient information about non-motorized road users such as pedestrians and cyclists. As a result, they are not adequately considered in traffic control and traffic safety. Even for motorized road users, high-quality information is often not available for centralized traffic control, resulting in inefficient traffic flow and unnecessary waiting times, for example at intersections or in traffic jams.

Research objective and relevance

The KIVI research project focuses on using artificial intelligence to control traffic and improve road safety in urban transport, which really includes all modes of travel. Through the use of AI processes, new control and safety functions are being developed, tested and applied in the traffic system of the city of Ingolstadt. The main objectives of the project are:

Increasing the efficiency of multimodal traffic control: By exploiting the previously untapped data potential across all road users, traffic control is to be optimized in order to improve traffic flow. This can be done, for example, through better traffic light control or electronic signage that directs traffic flows accordingly.
Increasing road safety: by applying AI methods, safety functions are to be developed that take into account all road users, including non-motorized road users, thereby increasing road safety. For example, maximum speeds can be set automatically so that motorized road users travel at slower speeds in the event of increased pedestrian traffic.

The High-Definition Test Field (HDT)

A central component of the research project is the High-Definition Test Field (HDT), which is used to research innovative mobility applications in urban areas. The HDT includes three busy traffic intersections in Ingolstadt as well as the expanded test site along Goethestraße. The choice of these nodes is based on several criteria, including early deployment of 5G mobile communications and the challenges posed by high traffic density and limited visibility, such as complex intersections and road layouts. The HDT enables the developed AI mobility applications to be tested in a real urban environment and their impact studied.

Research questions and approach

The project addresses several research questions in order to improve traffic flow and traffic safety and thus, for example, ensure lower emissions. These include:

Selection of suitable environmental sensors: Which sensors are best suited for installation in urban infrastructure to generate relevant information for optimizing traffic flow and increasing traffic safety?
Use of AI methods for state determination: How can AI methods be used to determine the state of dynamic objects based on raw data from stationary environment sensors? How are road users currently moving and what are the different movements of, for example, pedestrians, cyclists, car and motorcycle drivers, and heavy goods vehicles?
Fusion of environment sensor data with real-time requirements: Which AI techniques are suitable for fusing environment sensor data in real time to meet safety function requirements?
Efficient communication in the HDT: How can efficient communication within the HDT be ensured to meet real-time and robustness requirements? The latter refer to a system’s ability to avoid having to adapt its structure in the face of change.
Integration of the HDT into the existing transport network: How can communication between the HDT and the existing urban transport network be realized?
Warning concepts based on environmental sensors: Which warning concepts based on the stationary environment sensors in the infrastructure are most effective for the safety of vulnerable road users?
Potentials of traffic safety and flow optimization: What are the possibilities to contribute to increasing traffic safety and optimizing traffic flow with existing environmental sensors and advanced AI methods, and how can these potentials be determined on a real test field?

The answers to these questions are based on both existing data sources and newly collected data in the HDT. Special care is taken to ensure that the newly collected data can be seamlessly integrated into the existing backend systems and interfaces of the KIVI project partners.

The role of the project partners

The KIVI project is being carried out in cooperation between the Ingolstadt University of Technology (THI), AININ gGmbH and the Fraunhofer Application Center for Networked Mobility and Infrastructure. The partners are focused on the development and deployment of AI mobility applications, particularly in the area of road safety and traffic flow improvement. Close coordination with the city of Ingolstadt, where this is located, is crucial to ensure smooth integration of the research project into the existing infrastructure.

Ingolstadt University of Applied Sciences (THI) contributes its expertise in the fields of traffic systems and artificial intelligence. It is instrumental in the development of new control and safety functions based on AI processes. THI is working closely with the city of Ingolstadt to integrate the results of the project into the city’s traffic control system and thus contribute to more efficient traffic design.

AININ gGmbH focuses on the use of various data sources, including novel data sources, to accurately capture complex traffic situations. Through the use of AI techniques, innovative solutions are being developed to improve the transportation system.

The Fraunhofer Application Center for Networked Mobility and Infrastructure (FhG-AVMI) is involved in the areas of networked mobility and infrastructure. The center supports the research work in the KIVI project and contributes to the development and implementation of AI mobility applications. Through close cooperation with the other project partners, the aim is to develop innovative solutions for efficient traffic control and increased traffic safety.

Contribution to traffic safety and traffic flow optimization

The KIVI project aims to optimize the traffic system of the city of Ingolstadt by using artificial intelligence. By leveraging multiple data sources, including novel sensors and AI techniques, complex traffic conditions will be accurately captured and analyzed.

With the integration of AI mobility applications into existing infrastructure and communication between road users and infrastructure, traffic flow optimization and road safety will be improved. The HDT therefore provides a real-world test environment to try out the solutions developed and study their impact on traffic.

The project results will not only be relevant for the city of Ingolstadt, but can also be transferred to other urban transport systems. The KIVI project thus contributes to the development of future mobility systems based on artificial intelligence. This results in less congestion, less time lost in front of red lights but a more pleasant experience for pedestrians and cyclists in traffic. In addition, better traffic flow also offers an ecological advantage: reduced idle times reduce the consumption of vehicles with combustion engines, which also goes hand in hand with reduced pollution in the city.

Artificial Intelligence in the Ingolstadt Transportation System (KIVI)

Nikolai Zotov

Nikolai Zotov

Challenges in the current transport system

Research objective and relevance

The High-Definition Test Field (HDT)

Research questions and approach

The role of the project partners

Contribution to traffic safety and traffic flow optimization

More backgrounds