In our experiments, we will purposefully vary prices for both charging options. Your responses will be used to understand EV driver behavior around charging, based on concepts from behavioral economics and machine learning.

This data will be used to construct an optimal pricing and energy management system for the charging station.

PHASE I - Observe

How are we going to achieve our goal of re-designing EV charging services? We start as follows. In Phase I, SlrpEV stations operate just like any other charging station. Drive up, plug-in, start charging service. During Phase I, we will also:

  • Recruit EV drivers to join as research participants

  • Establish aggregate usage patterns

Phase I will occur for 3 months* after the station opens.

*This period is subject to change, depending on station usage, research participants recruitment, etc.

PHASE II - Learn

This is the phase where the most exciting research happens. EV drivers registered as research participants can select between two different charging service options:

  • "Immediate Charging", and

  • "Flexible/Eco Charging"

You can think about these options like Amazon delivery options: to enhance delivery logistics, different prices are offered for immediate package delivery or flexible package delivery. It’s that, but for EV charging and on a much shorter time-scale (don't worry charging should only take a couple of hours).

These two options will be accompanied by different prices in order to learn your choice behaviors. Your responses will be used to generate statistical models of EV charging behavior. These statistical models will be used to optimize pricing, to manage electric load and overstay, as described in Phase III.

Phase III - Optimize

In this last phase, we will optimize prices on charging menu, i.e, "Immediate Charging" and "Flexible/Eco Charging" using the data collected in Phase II. The optimization program minimizes costs while managing overstay. This will be done dynamically, i.e, each day, hour and vehicle arrival.

To get an understanding of how this will work, let's consider the example of 5 different cars with 2 of them opting for the Immediate charge versus the rest opting for Flexible charging. Immediate charging has an uncontrolled charging cost, since it depends upon the time of the day, whether it is peak hours or not. In contrast, the 3 EVs which chose "Flexible/Eco" charging, will have their charging optimized to avoid power peaks, while delivering the requested charge by their departure time.

As new EVs arrive, or when users need to change their input parameters, our algorithm will adapt to the circumstances to ensure that all the EVs are charged and the system's power limit doesn't overload.

What's Next?

So, now that you've seen how we're planning to make the next generation of EV charging possible, please join our research project! If you want to support SlrpEV, then share the website with friends and family! Also check out the wonderful people behind the project! Enjoy the photo gallery below, and follow us on Twitter (@slrpEV).