Research & Innovation Manager
A second-gen electrical engineer with a PhD in energy markets, Aleksei leads research into energy flexibility and market innovation.
He’s driving next-gen grid solutions to power a smarter, greener, and more resilient energy future.
Research Engineer
Arne brings a strong computer science background with a specialisation in AI to support a greener, more sustainable future. He earned his MSc in Computer Science, majoring in AI, from the Open University in the Netherlands. His master’s thesis, conducted in collaboration with Energised Futures, focused on the application of diffusion models on smart meter data.
Following his graduation Arne joined Energised Futures, continuing to explore innovative AI solutions for energy sustainability.
There’s no such thing as an “average” EV driver. When people plug in, how long they stay connected and how much energy they need can vary significantly. Understanding those differences is important as we design future tariffs, smart-charging services and energy systems.
SynthCharge is a generative AI model developed by Energised Futures to create realistic synthetic EV charging sessions. It learns relationships from real-world charging behaviour, including when vehicles are plugged in, how long they remain connected and how much energy is delivered, capturing patterns that occur across days and weeks.
Rather than creating one generic representation of an EV driver, SynthCharge can generate populations of charging sessions based on characteristics such as annual mileage, tariff type, battery capacity and vehicle class. This makes it possible to look at how charging behaviour and outcomes vary between different customer groups and scenarios.
Example EV charging sessions generated by SynthCharge over one month for 200 EV driver profiles across two groups: Battery Electric Vehicles on a time-of-use tariff and Plug-in Hybrid Electric Vehicles on a flat standard variable tariff.
To capture charging behaviour over time, SynthCharge represents each vehicle’s charging history as two timelines: when the vehicle is plugged in and how much energy is delivered, using 30-minute intervals. A conditional diffusion model then learns the patterns and relationships within this data, including recurring behaviours across days and weeks, to generate new charging sessions based on specified customer, vehicle and infrastructure characteristics.
By generating realistic populations of charging sessions, SynthCharge gives researchers and innovators a way to investigate EV charging without needing direct access to sensitive customer records. It can support tariff testing, grid-impact studies and analysis of how different customers could benefit from smart charging.
For a technical deep dive into how SynthCharge works, read our blog post Meet Synthcharge: Smarter Synthetic Data for EV Charging
Let’s accelerate it into a breakthrough, together.
Let’s collaborate.
