Secondary transformer stations – the points where medium voltage steps down to the voltage your power actually runs on – are some of the least visible parts of the grid. Higher voltage levels are wrapped in monitoring; these stations are too often heard from only after something's gone wrong: a protection trip, a field visit, a call from a customer with no power.
PANDORA is a research and development project building the visibility that's missing. It's designed to catch faults and equipment problems in secondary substations before they turn into outages – shifting maintenance from reactive to predictive, from “something failed” to “this needs attention, and here's why.”
PANDORA pairs standard sensor and telemetry data with fast, precise dPMU/PMU synchrophasor measurements of voltage and current. Machine learning and AI models process that data to recognize patterns, deviations, and early signs of equipment wear. The system uses explainable AI (XAI): it doesn't just raise a flag, it shows which data drove the call – so the decision stays with the person who knows the grid.
Results land in a clear interface: substation status, key measurements, flagged irregularities, risk scores, and recommended priorities for intervention. PANDORA is being built as a module within Thaora, ready to connect to the systems operators already run.
Secondary substations rarely get the same monitoring investment as the rest of the grid. That gap shows up as longer diagnostics, higher costs, and more time without power whenever something breaks – because nobody saw it coming.
Distributed generation, EVs, heat pumps, and battery storage are pushing more variability and two-way power flow through the grid, right as maintenance budgets and field teams are stretched thinner. Confirming that a fault happened isn't enough anymore – the value is in catching what leads up to it, early enough to act.
Distribution system operators – the teams monitoring the grid, maintaining equipment, planning investment, and managing technical data. Households, businesses, and public services benefit indirectly, through a more reliable supply with fewer and shorter interruptions.
Over the course of the project, PANDORA will be tested against concrete targets: up to 65% fewer unnecessary site visits, 40–50% faster repairs, 20–25% lower maintenance costs, and up to 70% fewer unplanned failures. These are project targets to be validated during development and pilot deployment, not guaranteed outcomes.
PANDORA brings several technologies together in a way built specifically for secondary substations – the part of the grid that's often the least monitored. It combines sensor data with high-precision dPMU measurements and AI models whose results are explainable to the people using them, and it isn't tied to a single equipment vendor, so it can be rolled out gradually, without replacing existing infrastructure.
PANDORA is developed by CompING computer engineering Ltd., the University of Zagreb Faculty of Electrical Engineering and Computing (FER), and Radio-Moto Ltd.
Comping leads the project as system architect, responsible for the platform, data infrastructure, analytics integration, API, user interface, security, and commercialization. FER is the scientific and methodology partner, developing the PMU/dPMU methodology and the AI/ML models. Radio-Moto is the field and integration partner, responsible for measurement and communication equipment, installation, and field testing.
Project name:
Development of a digital solution for predictive detection of operational disturbances and equipment failures in medium-to-low voltage transformer stations – PANDORA
Project holder:
CompING computer engineering Ltd.
Total project value:
EUR 2,827,715.31
Total eligible costs:
EUR 2,827,715.31
EU contribution:
EUR 1,936,280.86
Implementation period:
February 2026 – 31 January 2029
Contact person for further information:
Alojzije Jukić, CompING computer engineering Ltd.
This project is co-financed by the European Union through the Competitiveness and Cohesion Programme 2021–2027, funded by the European Regional Development Fund (ERDF).
The views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.
