AI for EV Charging: How Ampcontrol Built A Brain To Manage Smart Charging
As the transition to electric vehicles (EVs) picks up pace, there is a growing demand for technology to manage vehicle charging in a smart way.
In this article, we’ll answer four important questions:
- Why do electric vehicles need an AI brain?
- What is smart charging?
- What is Ampcontrol’s approach to smart charging?
- How can you implement AI for smart charging?
Why do electric vehicles need an AI brain?
The main reason that electric vehicles need an AI brain is to make charging faster and more efficient.
Electric vehicles are not the same as traditional internal combustion (IC) vehicles. Obviously, they are powered in different ways, but there is more to it than that.
The main difference is the speed with which you can refuel them. An IC vehicle can be fueled very quickly by topping it up with petrol or diesel. It’s not so simple with an EV that needs to be plugged into a charging point for a certain amount of time before it can be taken on a journey.
The amount of time that an EV needs to be charged for depends not only on the length of the planned journey but also on the efficiency of the charging process. In order to make EVs as efficient as possible, companies have invested a lot of time and effort into improving the speed of charging and designing larger batteries. This has led to significant improvements in how quickly you can charge vehicles for longer journeys.
However, there are some serious limitations on what can be achieved going forward, even over a period as long as the next 30 years, due to the chemical limitations of batteries, cost, and the physical constraints of EV charging.
The physical constraints include the installation and management of charging stations and the management of the charging service itself. One of the most important challenges to overcome when it comes to EV charging is how to optimise the process, which is the main reason why AI is needed for EV charging.
What is smart charging
Smart charging is a process that automates and optimizes decisions while an EV is charging. Optimized charging includes the following aims:
- Reduce total peak demand at the charging location
- Avoid high energy costs for both driver and charging point operator
- Ensure on-time departures and sufficient level of charge for each vehicle
- Help to stabilize real-time energy markets and electric utility services.
The first charging systems were built on very simple local microcontrollers and were useful for operating a small number of chargers per location or very basic optimization requirements.
This approach has become less feasible for most use cases, due to the following reasons:
- Expensive to install
- Lack of data (lacks latest values from the grid, vehicle, and pricing data.)
- Limited optimization objectives
- Local controllers don’t support AI-powered smart charging.
What is Ampcontrol’s approach to smart charging?
Our engineers at Ampcontrol have reinvented how people think about smart charging for electric vehicles. We saw past the limited optimization offered by microcontrollers.
Instead we identified several core needs for charging point operators, fleet operators, and energy companies:
At Ampcontrol we use AI to power reliable smart charging. Our algorithms cover all possible use cases (even non-realistic ones). The goal is that operators cannot make any configuration mistakes and possible hardware failures on the charger and vehicle side are eliminated.
We knew that charging would quickly become mainstream and that our customers would want a smart charging API that can be easily scaled up.
Hence, our engineer developed algorithms, system architecture, and interfaces that are easy to scale. Even before our launch in 2019 we simulated our algorithms with more than 100,000 historical charging events of customers and assuming more than 1,000 chargers per location.
Flexibility is related to the “smart” part of smart charging. Most use cases for smart charging require dynamic decision making, have multiple optimization objectives, and require data to be pulled in from various sources. Here at Ampcontrol, our customers often want to reduce energy costs, maintain on-time departure, and reduce total peak demand. Also, they want the whole thing to be automated. This requires a lot of flexibility.
Fleet managers want to know the answers to questions such as what happens if you:
- increase the number of chargers
- change energy rate
- add different vehicle types
- or enroll for utility demand response programs?
Ampcontrol provides a single tool that can answer all these questions and address them remotely, without hardware upgrades. It allows charging point operators to switch between optimization objectives and algorithms easily.
Ampcontrol is smart charging optimization software that can be integrated as easily as Stripe payment onto a website. The algorithms make the system self-managing and cover all possible edge cases.
Ampcontrol is accessible cloud-to-cloud, without additional hardware and agnostic for any type of OCPP backend or charge point manufacturer.
Companies like Revel are excited to work with Ampcontrol’s groundbreaking technology for smart charging.
How to get started with using AI for smart charging?
In recent months we’ve published several articles that guide you through the world of smart charging and Artificial Intelligence. Our goal is to lead you through the fundamentals for your decision process and development process.
Here are my favorites and our top-rated articles by readers:
- How to Send EV Charging Profiles to Your Open Charge Point Protocol (OCPP) Charging Station: Explains the concept and general format of OCPP and how to send charging profiles to chargers.
- 3 tricks to implement Smart Charging with OCPP 1.6: How to integrate Smart Charging with OCPP 1.6, rather than waiting for OCPP 2.0.
We’re also excited to demo Ampcontrol’s AI software and lead you through the integration process and roadmap for next year.
Reach out here: https://www.ampcontrol.io/demo