Amsterdam University of Applied Sciences

Big Data to create clean air

7 Oct 2014 17:15 | Communication

Amsterdam is leading the way internationally in the use of electric cars and wants to further expand the number of charging points in the city, as do other major Dutch municipalities. Jurjen Helmus of the HvA is carrying out research for all parties to determine how this can be done efficiently and effectively. Together with professor Robert van den Hoed he presents the CHIEF model, which brings together big data, and which can be used in cities throughout the world to roll out electric driving in the future.

Amsterdam is a polluted city, so the city authorities want to expand the use of electric cars. Amsterdam-Zuid already has the highest charging station density in the world, with an average of a 3-minute walk between charging points. “You have a very high chance of encountering an occupied charging point here,” says researcher Jurjen Helmus. The Amsterdam-Schiphol Airport taxi line has become fully electric, involving some 180 Teslas. Therefore, Amsterdam wants to expand by adding around 1000 new charging points. The Hague, Utrecht and Rotterdam also want to expand with another 1.5 times as many charging points.

Dilemmas

The Netherlands’ largest municipalities (the G4) want to roll out the infrastructure for the charging points. Jurjen Helmus: “None of the parties involved – the municipal authorities, the grid operators and the electricity suppliers – have any idea what the business case looks like yet or how they need to approach it.” The loading points are currently paid for by the municipal authorities, at a cost of around 6000 to 8000 euros each, but power company Essent/Nuon is making a profit. Who is going to pay for the new charging points? Where should they be located? And will each point have its own fuse box, or will we work with extension cables? Do we want rapid charging points or slow charging points? Jurjen Helmus is going to work out the desired model for all parties, by analysing big data on the loading points and the users.

Figure: the yellow areas in Amsterdam-Zuid still have room for more loading points

Data are the new gold

The Amsterdam University of Applied Sciences will have exclusive access to data on Amsterdam’s charging infrastructure. “As an objective research group we are a safe party for the analysis of these data. Market parties do not have access to these data,” says researcher Jurjen Helmus. “Data are the new gold.” The researchers will gather anonymous data for each charging pass ID on charging time, the type of car and charging point usage. “First we will clarify the business case, and then we will develop a model that predicts charging behaviour and usage. This model, called CHIEF (Charge Infrastructure Efficiency Model), can be used to roll out electric driving in any city in the world.” The researchers recently gave a presentation on their method to officials in San Francisco. 

CHIEF predicts

The CHIEF study being conducted by the Urban Technology research group has the potential to expand from Amsterdam to the entire Randstad conurbation. The CHIEF study has four phases. First the KPIs for all stakeholders were identified. Now the researchers are collecting new data monthly, and will do so daily in the future. These data are then processed in a data pool.

In the second phase the researchers will investigate the reasons for charging behaviour and charging point usage. They will use cluster algorithms to look for patterns in the data and identify the predictive factors. The third phase will consist of forecasting with those predictive factors. And then it will be time for the final step: simulations. Jurjen: “We will make a simulation model of the charging infrastructure, a digital miniature version of Amsterdam. CHIEF is a reliable predictive model.”

 

Students are involved in the research too. The statistics courses are already working on the database, and Technology students will take data engineering as an additional subject. Students are using data mining to studying relationships between the variables, and more than 10 graduates are conducting research as part of the CHIEF study.