Stadtwerke Gießen is preparing to revolutionise local public transport with a pioneering research project. The initial results are already available.
When do how many people travel on which bus and where exactly is the bus at any given time? The aim of the NV-ProVi project, which is funded by the Federal Ministry of Transport and Digital Infrastructure and on which Stadtwerke Gießen (SWG) and the consultancy firm Brodtmann Consulting are currently working, is to be able to answer these questions precisely at any time. In fact, this information is becoming increasingly important. After all, it serves to noticeably optimise local public transport services - both in terms of attractiveness and cost-effectiveness. "Our biomethane-powered, virtuallyCO2-neutral city buses are an essential part of the transport transition that we are consistently and vigorously driving forward here in Giessen. The NV-ProVi project is a decisive step forward for us," says Jens Schmidt, Commercial Director of SWG, explaining the background.
What exactly is it about?
NV-ProVi is a genuine innovation: vehicles from SWG's local transport subsidiary MIT.BUS send data on their position and occupancy to a server. Here, artificial intelligence (AI) calculates a forecast for the utilisation of the buses and makes it available. All of this happens in real time. "Our data is of interest to many people," says Jens Schmidt. He names national transport organisations, companies that process public transport data - for example to develop new services - and universities that can make good use of such detailed information for research purposes, such as the creation of reliable transport models, as potential customers. Prof Dr Jörg Pfister from the Technische Hochschule Mittelhessen puts it like this: "Empirical research thrives on up-to-date, freely available data, which is often difficult to obtain. That's why projects like NV-ProVi are real gold mines for us."
Last but not least, those responsible in the city of Giessen should also be paying close attention to the subject matter. "The findings will help us to improve local transport services together with SWG. And thus convince more and more people to switch to the bus. As a result, NV-ProVi is contributing to our goal of making the city climate-neutral by 2035," summarises Gerda Weigel-Greilich.
Like many others, the department head responsible for transport is delighted with the project's progress. A lot has already happened since the start in March 2020. For example, MIT.BUS has installed all the passenger counting systems financed by the grant. Since March of this year, all vehicles have been constantly providing real-time data on their position. 25 of the 56 buses also send the updated number of passengers after each departure at a stop. Software specialists have developed the first AI prototypes required for further processing, which they are currently optimising for precision in forecasting and the speed of the algorithm.
First visualisation
A so-called LiveMap was created to visualise the interim results. This visualises the information on position and occupancy available in real time. What is special about it is that, unlike most systems currently in use, the AI does not interpolate the positions of the vehicles between stops. Instead, it processes GPS data, which is updated every ten seconds at the latest. As a result, the AI delivers meaningful results even if buses take a different route than usual due to unplanned diversions, roadworks or other traffic obstructions. "The NV-ProVi-LiveMap can easily adapt to the current situation and always shows correct information," summarises Mathias Carl, Managing Director of MIT.BUS.
A clever mix is used for the capacity utilisation displays. For buses equipped with passenger counting systems, the AI uses their real-time data to display the current values. It calculates the forecasts for the rest of the journey from historical information. This data, which has been collected over the years, is also used for vehicles that do not have a counting system on board. Here, the AI analyses the overall public transport situation in Giessen and creates forecasts for all vehicles.
"However, the current LiveMap of Giessen is not planned as a permanent installation," Jens Schmidt points out. Rather, the data will sooner or later be incorporated into the LiveMap of the Rhine-Main Transport Association, which has been actively supporting the project for two years. Nevertheless, the existing visualisation is suitable for conveying where the journey is going and what passengers can expect. In fact, it should soon be possible to see on a smartphone in real time exactly when the bus will arrive, how full it is and whether it still has space for a pushchair. Information that shows alternatives with free seats or at least significantly lower occupancy is also conceivable. In short: NV-ProVi is primarily aimed at making it easier for passengers to use buses and trains. With quick and easily accessible information. One of these features is already available in a somewhat reduced form: Since the beginning of May, SWG have been showing how full the buses are in a clear graphic. For every route, at any time and for every section of the route.
Technologically at the forefront
With its commitment to NV-ProVi, SWG is once again proving to be a driver of technical progress. "We are working on one of the most innovative public transport projects in Germany," explains Jens Schmidt. And Gerda Weigel-Greilich adds: "The city of Giessen promotes innovation and is proud to be a technological pioneer together with SWG."
Keyword promote: NV-ProVi has what it takes to decisively advance public transport. Consequently, the results of the research project are also of great interest to other local authorities and transport service providers. This is the reason why the federal government is subsidising the project with more than 350,000 euros. The funds come from the mFUND research initiative and are only awarded after rigorous scrutiny.
Prof Dr Jörg Pfister can assess how useful NV-ProVi is for the traffic turnaround: "We use the data to realistically simulate traffic in Giessen using detailed models. With this approach, good ideas for improving traffic flows can be implemented virtually and meaningful information on the expected impact can be generated. This information then forms the basis for making the right decisions for sustainable mobility in Giessen and elsewhere. Of course, this approach can be transferred to any municipality - if the relevant data is available."