Decision Support for Roadwork Coordination

With increasing urbanization the cities are growing and this puts pressure on the infrastructure. The Decision Support for Roadwork Coordination is developed to simplify the coordination of roadwork and maintenance.

The challenge 

As cities grow and the demand for mobility services is increasing, traffic is becoming a major challenge for a lot of cities. To deal with this, traffic infrastructure has to be maintained and expanded to meet the needs of people. Unfortunately, this requires construction sites which restrain the traffic flow.

Therefore, there is a big necessity to plan and coordinate the roadworks. In the last years different tools such as ROADS or NyStart have been developed to handle this challenge. These tools support the experts in their decision processes of coordinating roadworks, for instance the selection of roadworks and their timing. However, the past years have proved that the experts make decisions based on their individual experience. This results in a need for more experts or training and qualifying new employees that are able to fulfill the challenging task of roadwork coordination. 

The solution 

In this solution different machine learning approaches and specifically deep leaning methods will be applied on the challenge of roadwork coordination. Decision rules will be developed by coordinating experts from different areas who are directly involved in decision making for construction work and thereby a decision support system (DSS) will be developed. Such a DSS will support the experts in their complex work with roadwork coordination. The system aims to result in more harmonized decisions. Moreover, the training of new staff members is expected to be easier and in the end the quality of the decisions may increase.

The Decision Support for Roadwork Coordination solution is at its very first stage. 

The team

Hamburg and Gothenburg face the same complex challenge of roadwork coordination and they are addressing the challenge together with the University of Bradford. The University of Bradford is supporting the cities with expertise in machine learning approaches and specifically in deep learning methods. Hamburg and Bradford University will lead the way and Gothenburg will follow up later.

Why is your solution a good fit for the SCORE project?

Developing a tool that can be implemented in different roadwork coordination software of different cities demands high regard for standardization and openness in data and coding. As these are the fundamentals of SCORE, this solution fits very well in the project's scope.

The ambition is to develop a decision support tool that can be implemented in different roadwork coordination tools.


Thorben Finger: