Adaptive Charging Network

Adaptive Charging Network Research Portal

PI: Steven Low (Division of Engineering and Applied Science)

The historic transformation of our energy systems points to a future where there are billions of distributed energy resources (DERs) throughout our energy infrastructure. The development of an intellectual basis and practical algorithms for the planning, design, and operation of such a large network of DERs is one of the most important challenges facing our energy systems transformation.

Electric vehicles (EVs) will be a critical DER in future smart grids. Working with Caltech Facilities and PowerFlex, a Caltech startup, the research group of Steven Low has designed, developed and deployed an Adaptive Charging Network (ACN) for electric vehicles (EVs) at one of Caltech’s parking garages. ACN is capable of real-time measurement, communication, computation and actuation. It has been operational 2+ years, having delivered 2M electric miles and avoided almost 700 metric tons of CO2e.


The Caltech ACN

The mathematical or simulation models used in algorithm research for smart grid are often unrealistic compared to the real system for which an algorithm is designed. By abstracting away many complicating details, mathematical analysis allows us to uncover hidden structures of the problem, leading to new qualitative insights and new algorithmic designs. There is however a sizable gap between theoretical results and a robust and high-performance implementation of an algorithm. Bridging this gap is critical to making an impact in practice. It will require, beyond mathematical analys, detailed simulations driven by realistic models and data the ability to test an algorithm implementation in the field. The lack of these capabilities is the bottleneck in crossing this gap, and the Schmidt Academy is working with the Low group to develop a suite of software, called the ACN Research Portal, that will make Caltech ACN not just a production facility, but also but also a live testbed for algorithm researchers at, and outside, Caltech.

As part of this collaboration between the Low Netlab research group and the Schmidt Academy, we refined and completed a suite of software which was fully documented within a system of continuous integration with testing. We also collaborated with the group on two research projects one involving battery modeling and the other on modelling the energy flow associated with bus charging depots. The three components of the software suite are:

ACN Data: provides EV charging data to research community
ACN Sim: simulates EV charging
ACN Live: experiments with real EV charging

The ACN Software
Github link: