IN Brief:
- The deployment expands a pilot of more than 400 public chargepoints to 1,500.
- OCPI 2.2.1 Charging Profiles allow optimised schedules to be sent directly to chargepoints without a vehicle API.
- The chargepoint operator retains control of the infrastructure while participating drivers can shift charging away from peak demand.
Ubitricity and Deftpower are expanding interoperable smart charging to 1,500 public chargepoints in the Netherlands using OCPI 2.2.1 Charging Profiles.
The deployment follows a pilot covering more than 400 chargepoints and changes how an optimised charging instruction reaches the vehicle infrastructure. A mobility service provider can send a schedule directly to a compatible chargepoint without relying on a separate API connection to the car.
The chargepoint operator retains control of sessions on its infrastructure, while the optimisation can incorporate preferences set through the driver’s charging service. That architecture broadens the pool of vehicles that can take part because participation is not limited to cars exposing a compatible remote-control interface.
Ubitricity operates more than 20,000 public chargepoints in the Shell Recharge network across the Netherlands, UK, Germany, and France. The Dutch programme therefore remains a limited part of its estate, but 1,500 live public points provide a significantly larger interoperability test than the initial pilot.
Charging control shifts towards infrastructure
OCPI is used to exchange information between chargepoint operators and e-mobility service providers. The Charging Profiles functionality adds the ability for an optimised charging schedule to pass through that relationship and be executed by the charging infrastructure itself.
That avoids one source of fragmentation in managed EV charging. Vehicle manufacturers provide different levels of remote access, and not every model exposes the same control interfaces, leaving software providers dependent on brand-specific integrations when optimisation sits entirely on the vehicle side.
Moving the schedule to the chargepoint does not eliminate compatibility requirements. The chargepoint, operator platform, mobility provider, and roaming interfaces still have to support the relevant protocol functions consistently, and legacy equipment can limit how uniformly the model is deployed across an established network.
The benefit appears when many modest charging loads are aggregated. One AC session has little effect on a national power system, but thousands beginning at similar times can reinforce local evening peaks and increase loading on distribution assets.
Shifting a proportion of those sessions later reduces coincidence without necessarily reducing the amount of energy delivered to drivers. Deftpower says 65% of users who adopt smart charging leave it activated by default, indicating that automated scheduling can remain in use after the first session rather than requiring repeated intervention.
Drivers using participating services can also receive cashback for moving demand away from peak periods. That creates a direct incentive at the point where the timing decision is made while leaving the infrastructure operator responsible for the technical limits of the charging estate.
Flexibility cannot replace network capacity
The Netherlands provides a demanding environment for the model because grid congestion is already limiting new electrical connections in several areas. Managed charging can improve utilisation of an existing connection, but it cannot create transformer, cable, or substation capacity where a sustained physical constraint has already been reached.
The distinction becomes more important as charging networks expand. European public charging capacity already exceeds fleet-based AFIR requirements across most member states, but aggregate charger numbers conceal major differences in location, power rating, utilisation, and local network availability.
Public AC charging is particularly suited to flexibility where vehicles remain connected for longer than they need to draw power continuously. A vehicle parked for several hours provides a time window within which charging can be moved, unlike a rapid-charging stop where the driver’s principal requirement is to restore range as quickly as possible.
For chargepoint operators, the control boundary remains important. They still have to manage site capacity, hardware availability, faults, metering, session data, and any limits imposed by the distribution network even when a mobility provider sends the proposed charging profile.
Scaling open-protocol control across several operators would reduce the need for bespoke integrations, but consistent implementation will determine whether the theoretical interoperability survives contact with different back-office systems and generations of charging hardware.
The 1,500-point deployment gives Ubitricity and Deftpower enough live infrastructure to expose those differences. Successful scheduling has to work repeatedly across normal public charging sessions rather than under the controlled conditions of a small technical demonstration.
EV charging will continue to require physical network reinforcement as vehicle numbers and high-power infrastructure grow. The Dutch programme addresses a narrower engineering question: how much additional utilisation can be extracted from connections already in place by controlling the timing of flexible demand more precisely.
If OCPI Charging Profiles perform reliably across the expanded estate, the result will be a practical route for public chargers to participate in demand flexibility without requiring vehicle-specific control integrations. The next constraint will then be less about whether the protocol works and more about how widely operators and mobility providers implement it.
