EV charging stations become cheaper with AI - mostly

It is quite a challenge to find out the prices, charging capacity and any discounts at charging stations at certain times of the day and night, especially when you are on the move.

After all, the providers want to earn something, but they also pass on the fluctuating electricity prices. It is also worthwhile if the charging station is well utilized. If you are looking for a charging station, you want to charge as cheaply as possible, not have to wait and, ideally, get a high charging capacity for a short standing time.

A test arrangement using artificial intelligence was able to satisfy both sides equally. A specially trained AI was allowed to submit its own offers for a large number of charging stations and thus decide how high the charging fees should be.

To do this, it had to be able to retrieve information on the location and current battery status of the respective e-cars. So you have to be prepared to disclose quite essential data. From a financial perspective, however, it could be worth it.

In almost all cases, the AI offer was cheaper than the charging station operator's normal tariff. This is also worthwhile for the other side, because better utilization of the charging stations is achieved through price-controlled distribution. In addition, charging can be made more attractive at times when electricity can be purchased very cheaply. A kilowatt hour only costs a few cents on the electricity exchange, especially during sunny midday hours.

However, there was one exception in the experiment: if the battery's charge level meant that only a few or even a single charging station was available, the price that the AI considered appropriate skyrocketed. It imitated human behavior a little too well.