A quantum-assisted optimisation study focused on peer-to-peer EV charging networks has been positively evaluated under the EuroHPC Quantum Access Pilot Call. The application, led through Negentra Yazılım ve Oyun Teknolojileri A.Ş. by the team behind Veevex, has been awarded 10 QPU hours on Piast-Q, a 20-qubit ion-trap quantum system, for a three-month access period.
The research explores whether constrained EV charging assignment and scheduling problems can be represented as Quadratic Unconstrained Binary Optimisation (QUBO) models and evaluated with the Quantum Approximate Optimisation Algorithm (QAOA) on real quantum hardware.
From validated simulation to physical hardware
The project already includes a validated 9-variable QUBO proof of concept, exhaustive verification across all 512 binary states, exact classical benchmarking and ideal-statevector QAOA experiments at depths p=1–3. The next step is physical execution on Piast-Q.
The hardware experiments are designed to measure effects that ideal simulation cannot establish reliably, including sampling behaviour, hardware noise, feasible-solution recovery, run-to-run variance, circuit execution characteristics and scaling toward 16- and 20-variable marketplace instances.
What the EuroHPC technical assessment highlighted
The EuroHPC technical assessment accepted the proposal as technically feasible and confirmed that no additional emulator benchmarking is required before physical-hardware access.
- The reviewer described the hardware-readiness preparation as “exemplary”.
- The assessment noted that this was the only proposal in that cut-off to have integrated with the site tooling in advance.
- The existing Qiskit-based RZ/RZZ cost-unitary workflow and RX mixers were assessed as suitable for the selected physical hardware.
- The full requested allocation of 600 QPU minutes (10 hours) was awarded for the three-month project period.
Technical assessment
The final classification cited a fully verified prototype, native-aware circuit design, pre-built integration with the system’s actual software stack and a minimal, realistic resource request.
A deliberately evidence-led quantum study
The research is not designed to claim quantum advantage at the current hardware scale. The objective is to establish experimentally grounded evidence about where quantum-assisted optimisation may or may not become useful for distributed EV charging and energy coordination.
The study will compare quantum-derived solutions with exact or mathematical classical baselines and evaluate feasible-solution probability, optimum and near-optimum recovery, objective-value gap, sampling stability, run-to-run variance, transpiled circuit depth and gate counts.
For the planned 16- and 20-variable experiments, the reviewer specifically recommended prioritising shallow QAOA depths and checking compiled two-qubit gate counts against the available hardware depth budget. The assessment also encouraged repeated runs across different calibration periods to strengthen the variance analysis.
Why this matters for Veevex
Veevex is building a private AC charger sharing network in which drivers, charger owners and site rules must be coordinated under real operational constraints. This research provides a controlled environment for testing advanced optimisation methods against one of the core computational challenges behind distributed charging networks: matching demand with available charging capacity while respecting constraints.
The immediate goal is research validation rather than production deployment. Findings from the Piast-Q experiments will help determine whether further hybrid quantum-classical development is technically justified for larger peer-to-peer charging marketplaces.