
Episode 60 - Zero Downtime and Safer CPMS Migrations: Why CPOs Need Broker Forking
Summary Most conversations about EV charging infrastructure focus on chargers, power and grid connections. Far fewer look at backend architecture, even though it decides whether a site keeps charging when something goes wrong. In the traditional setup, each charger talks to a single charge point management system (CPMS) over OCPP. If that one connection drops, transactions, remote commands and monitoring can stop with it. For critical sites like electric bus depots, and for CPOs migrating networks after an acquisition, that single point of failure is a real business risk. In this episode of the Charge Point Podcast, Christophe Lephilibert speaks with two guests from inside FLEXECHARGE. Szymon Żmijewski is the software developer who built broker forking, FLEXECHARGE's OCPP proxy feature for backend redundancy, designed in close collaboration with Daimler Buses for bus depot operations. Jan Köster, CCO and co-founder of FLEXECHARGE, brings the commercial view from six years of working with CPOs on load management and energy management. Together, they explain how broker forking lets one OCPP proxy maintain connections to a primary, a fallback and a service backend at the same time, why preserving transaction state matters more than simply reconnecting, and how the same architecture lets CPOs migrate between CPMS platforms gradually instead of in one risky switch. They also zoom out to CPO consolidation and specialisation across Europe, and Jan shares the three industry trends he is watching most closely. Chapters What you'll learn Why the traditional one-charger, one-backend model creates a single point of failure that can paralyse an entire charging site when the internet connection drops. How broker forking routes OCPP traffic to three backends, a primary, a secondary fallback and a service backend, each with a distinct role. How a fallback backend keeps charging sessions running at a bus depot when the connection to the primary CPMS is lost. Why OCPP 1.6 transaction IDs make multi-backend operation hard, and how a mapping layer in the broker keeps three backends in sync without the charger noticing. How the service backend separates monitoring and diagnostics from operations, and how real-world use led to a whitelist of messages it is allowed to send. How broker forking turns a CPMS migration after an acquisition into a controlled, gradual move with sandbox testing. Why Jan sees three trends shaping the market: the shift from hardware to controllable energy assets, the specialisation of CPOs and their tech stacks, and standardisation as the key to winning the consolidation game. Why an OCPP proxy with forking can match the reliability of a direct Modbus integration with the charger. Notable quotes "The primary purpose of a charging site is not to be connected to the internet, it's to charge cars. So when the internet connection fails and that paralyzes the whole site, that creates a serious problem." - Szymon Żmijewski, FLEXECHARGE "The charger thinks that there's only one ID, each of the backends think that there's only one ID, but we internally know that actually there's three different IDs." - Szymon Żmijewski, FLEXECHARGE "The CPOs have realized if they handle each charging site as an individual project, there's no chance to scale. And the ones that have figured it out to standardize are the ones that are taking the lead in the consolidation game." - Jan Köster, FLEXECHARGE About the guests Szymon Żmijewski is a software developer at FLEXECHARGE, the European EV charging software and hardware company, where he works on energy management solutions and hardware integrations for charge point operators. He is the developer behind broker forking, the OCPP proxy feature that gives CPOs backend redundancy, and has worked hands-on on the state preservation and transaction ID mapping that make multi-backend OCPP setups reliable in real-world fleet environments. Jan Köster is CCO and co-founder of FLEXECHARGE, where he leads sales, business development and account management. He has helped grow FLEXECHARGE from load management for small customers into a comprehensive energy management platform for CPOs, with features like broker forking. With a commercial background in industrial control systems for heating, ventilation and lighting, he brings a practical view of the reliability and scalability that critical charging infrastructure demands. Szymon Żmijewski on LinkedIn: https://www.linkedin.com/in/szymon-zmijewski/ Jan Köster on LinkedIn: https://www.linkedin.com/in/jankoester/ Related listening: Episode 13 with FLEXECHARGE CEO and co-founder Max Brandt, on how the company evolved into a vendor-independent grid-integration platform. Got some feedback or questions? Text us! You can get more information on our website. Give us some stars please 🌟 If you enjoyed the episode, please give it a 5-stars note and leave us a review. Get in Touch: Do you have suggestions for future guests or topics? Reach out to us at hello@flexecharge.com. Subscribe & Stay Updated: 🔔 Make sure to subscribe to the Charge Point Podcast to stay updated on the latest trends and developments in the world of EV charging and energy. 📩 You can subscribe to our Newsletter, The Charge Point Journal, by clicking here. 🫱🏻🫲🏼 And you can follow us on Linkedin by clicking here and connect with Christophe here. ⚡ Let’s keep the conversation—and the electrons—flowing.


















