Steinbeis experts use AI methods to analyze travel purposes
In the German federal trunk route network automatic traffic counting stations are installed which permanently record the number of vehicles passing certain intersections. These counts result in repeating demand profiles, which are the basis for dimensioning the infrastructure. It is not known, however, what the motivations i.e. trip purposes are behind these demand profiles. As part of a project called NN-FAND, experts at the Institute of Transport Studies at Karlsruhe Institute of Technology (KIT) together with STASA Steinbeis Applied Systems Analysis conduct a feasibility study on possibilities to merge information on the temporal demand profiles at counting stations with other contextual information and data. The idea is to gain insights into the demand structures in terms of the different travel purposes which superpose to the measured profiles in the highway network for allowing an improved understanding for dimensioning the road infrastructure.
The Federal Ministry of Transport and Digital Infrastructure (BMDV) uses permanent traffic counting systems to automatically and continuously monitor the volume of vehicles passing through the intersections of the federal highway network. The information collected is used to dimension the road infrastructure, e.g. the number of lanes required to serve existing demand – efficiently and without disrupting traffic.
However, by reasons of technological, demographic and societal processes the demand volumes are likely to change at least in the split up of different travel purposes. And this is likely to affect the dimensioning as well.
NN-FAND provides more and better data
Although continuous automated counting offers insights into the timing and progression of traffic volumes in the form of demand profiles, there exists until now only rough information on the breakdown of that profile to different trip purposes or the user groups. The aim of the NN-FAND project is therefore to come up with simple estimates, to derive information from the demand profiles who is travelling and why. Standardized travel behaviour surveys such as the German Mobility Panel conducted by the German Minister of Transport and Digital Infrastructure (BMDV) contain information on the trips performed by individuals by mode, purpose and timing however without spatial reference.
Demographic and societal trends have a direct influence on travel purposes. For example, the share of the population working is likely to decline due to demographic changes. In addition, physical means of transportation are increasingly being replaced by “virtual travel,” for example due to people working from home. “This results in different patterns of travel motivation, especially during periods of peak demand, which are relevant for infrastructure dimensioning,” explains Professor Dr. Günter Haag, Managing Director of STASA.
Having examined the data from the mobility surveys, the project team has drafted “ideal” trip purpose differentiated demand profiles to provide usage patterns based on defined characteristics. These are being used to reproduce the demand profiles for individual fixed counting stations. For example, AI methods are being used to transfer usage patterns to other counting station data by looking at the similarities between infrastructure and environmental factors (type of location, position within the network). The aim is to gain a better understanding of usage patterns (trip motivations, users) and arrive at a causally justifiable prediction of the robustness of measurements under different circumstances.
To conduct its evaluations and merge information from different sources, the team of experts at STASA is using innovative AI methods, with the ultimate intention of demonstrating the specific applicability of those methods to the feasibility study.
As part of the mFUND innovation initiative, since 2016 the Federal Ministry of Transport and Digital Infrastructure (BMDV) has been funding data-centric R&D projects with a bearing on digital and connected travel, or so-called Mobility 4.0. Project funding is underpinned by active professional networking between stakeholders in politics, business, public administration, and research, as well as the provision of open data on the mCLOUD portal.
For more information, go to www.mFUND.de.