Use cases

AI-powered grid solutions for TSOs and DSOs

Comprehensive solutions for transmission and distribution system operators, powered by reinforcement learning and digital twin technology. Every module runs on the same digital twin — the twin you build for one module carries over to the next.

ONE DIGITAL TWIN · FOUR MODULESMV gridBESS · power-to-heat2connection point1PV feed-inswitchgear431Operating envelopes2Flexible asset management3State estimation4Congestion management
Where each module acts on the feeder — all four run on the same digital twin.
1
Discharges: ElWG flexible network access + capacity publication

Dynamic Operating Envelopes & Hosting Capacity

Flexible network access is now Austrian law (§ 103 ElWG): under the temporary route, full access is due 12 to 24 months after contract conclusion, depending on network level. Dynamic operating envelopes make flexible access workable — static or dynamic connection limits that follow the actual state of the grid instead of the worst case. FLEX computes them per connection point and keeps the published hosting-capacity picture current.

99.99% of operational voltage violations mitigated on an Italian DSO grid (ENEL), computed ~5,000× faster than the baseline
Operating envelopes in depth →
2
Discharges: 2027 peak-capping — minimise real curtailment

Flexible Asset Management & Remedial Action Curation

From 2027 the ElWG caps PV feed-in at 70% of module peak — without compensation. The operator's job flips from granting access to minimising what actually gets clipped. FLEX schedules batteries, power-to-heat and other controllable assets against forecast grid states, and automatically generates and stress-tests the remedial action lists your control room reaches for when reality deviates from plan.

BESS and power-to-heat in scope as controllable assets
Flexible asset management in depth →
3
Discharges: the observability every other duty depends on

Grid State Estimation

You cannot optimise a grid you cannot see. Most distribution grids are observable only in patches — smart meter data arrives late, sensors are sparse, and forecasts rarely reconcile with actuals. FLEX fuses the measurements you already have into a physically consistent real-time picture of the MV grid, accurate enough to plan against. Measured with Stadtwerke Kapfenberg: estimation error reduced from ~8 kW to below 5 kW.

~8 kW → <5 kW estimation error, measured
State estimation in depth →
4
Discharges: delivering ElWG deadlines without unlimited construction budget

Congestion Management & Topology Optimisation

Redispatch and curtailment cost money and goodwill; building takes a decade. Between the two sits a lever most grids barely use: their own switchgear. FLEX searches the enormous space of feasible switching states and surfaces the few that measurably lower loading on stressed lines — validated against the full physics of your grid. Measured with Stadtwerke Kapfenberg: peak line loading down from 60.2% to 42.1% — 18.1 percentage points — without a single new asset.

60.2% → 42.1% peak loading, measured on a real Austrian grid
Congestion management in depth →
Use cases

Every module runs on the same digital twin

The twin you build for one module carries over to the next.

Where to start

Most operators start from a concrete operational problem rather than a module. We wrote a page organised that way.