Live grid console • simulated data

Renewables are variable. The grid can’t be.

HELIOS is Ace Hacker's grid-intelligence engine. It forecasts every megawatt of sun and wind, schedules storage and dispatch with quantum optimization, and holds the grid stable in real time, so operators can push renewable penetration toward 100% without losing the lights.

Request a demo → Explore the engine

50.00 Hzfrequency held ±0.02

312×faster dispatch solve

3.1%day-ahead forecast MAPE

helios://grid/live-balance BALANCED

Generation mix Frequency Price

System load

34.2 GW−1.2%

Renewable share

68%

Battery SoC

72%chg

Engine at a glance

Renewable penetration sustained in pilot

Forecast + Optimize

Day-ahead generation forecast MAPE

Forecast

0 GWh

Curtailment avoided per day

Optimize

0 Mt

CO₂ avoided per year, per region

End to end

One engine, four instruments

The whole balancing problem, from forecast to frequency.

HELIOS pairs each part of grid operation with the right computation, deep learning to predict a chaotic sky, quantum optimization to schedule a combinatorial grid, and real-time control to keep it all stable.

Forecast

Generation & demand

Weather-to-power models forecast solar and wind and load hours to days ahead, with calibrated uncertainty.

spatio-temporal • probabilistic

Optimize

Dispatch & storage

Quantum and hybrid solvers schedule generation and storage across a combinatorial space, at cost.

QAOA • unit commitment

Stabilize

Real-time stability

Frequency and voltage control plus contingency screening keep the grid inside its limits, continuously.

power flow • N-1 • control

Materials

Storage discovery

Quantum chemistry screens electrolytes and electrodes for the next generation of grid-scale storage.

VQE • formation energy

Forecast • Instrument 01

You can’t schedule what you can’t see coming.

A grid running on sun and wind lives or dies by its forecast. HELIOS turns weather into power: it predicts generation and demand from hours to days ahead, and it tells you how sure it is, so the optimizer can plan for the whole distribution, not a single guess.

01
Weather-to-power modelsNumerical weather ensembles drive learned models that map irradiance and wind fields to plant-level output.
02
Probabilistic by defaultEvery forecast is a distribution (P10–P90), not a point, because a grid plans against the tails, not the mean.
03
Spatio-temporalAttention over the network captures how a cloud front or a wind ramp moves across assets in space and time.

Graph + temporal attentionQuantile lossNWP ensemblesSCADA

Solar generation forecastnext 24h

power (MW) vs hour

Forecast P50P10–P90 bandActual

Wind resource • forecast mapm/s by hour × site

Forecast skill by horizonlower MAPE better

Optimize • Instrument 02

Scheduling a grid is combinatorial. We solve it in time.

Unit commitment, deciding which generators and batteries run when, is a mixed-integer problem that grows explosively with assets and constraints. HELIOS maps it to a quantum solver and returns a least-cost dispatch that respects every ramp, reserve, and network limit, fast enough to re-solve as the forecast moves.

Optimal dispatch stack24h • least cost

generation by source (GW)

SolarWindHydroStorageGas peaker

01
QAOA on the commitment problemOn/off decisions are encoded into a cost Hamiltonian with ramp, reserve, and network constraints as penalties.
02
Storage co-optimizedBatteries are scheduled jointly with generation, arbitraging the forecast to soak up cheap renewables.
03
Warm-started & hybridA classical relaxation seeds the solver, so it converges within the operating window.

QAOA / QUBOMILP baselineCVaR objectiveDC-OPF constraints

Solver cost convergenceobjective per iteration

Battery state of chargecharge / discharge

SoC %

Stabilize • Instrument 03

Between the plan and reality is milliseconds.

A dispatch is a plan; the grid is physics. As inverter-based renewables replace spinning mass, inertia falls and frequency moves faster. HELIOS watches power flow in real time, screens every credible contingency, and acts before a disturbance becomes an outage.

System frequency50 Hz ± band

50.00 Hz

FrequencyDisturbance + responseOperating band

Live power-grid flownodes & line loading

Contingency (N-1) screening • line loading under each outage% of thermal limit

Materials • Instrument 04

The transition needs storage that doesn’t exist yet.

Cheaper, denser, longer-lived storage is the missing piece of a renewable grid, and it comes down to chemistry. HELIOS screens candidate electrode and electrolyte materials with quantum chemistry, computing the electronic properties that classical approximations get wrong.

01
Formation & voltage from first principlesVQE computes formation energies and redox potentials for candidate chemistries where DFT is unreliable.
02
Multi-property screeningEnergy density, cycle stability, and ionic conductivity are estimated together to rank real candidates.
03
Beyond lithiumSodium, solid-state, and flow chemistries are all in scope, the grid does not need to fit in a phone.

VQEActive-space embeddingNa-ion / solid-statehigh-throughput

Candidate storage materialsscreened library

energy density × cycle life

CandidatesPareto-optimalSelected for synthesis

VQE formation-energy convergenceHartree

Cell voltage by candidatevolts

Benchmarks • honest numbers

Where quantum earns its place, and where it does not.

We benchmark against strong classical baselines, tuned MILP solvers, state-of-the-art forecasters. Quantum optimization wins as the commitment problem grows; classical methods still win on small cases, and HELIOS routes accordingly.

Dispatch solve time vs problem sizelower better

Classical MILPHELIOS hybrid

Capability profilevs incumbent EMS

Architecture • from telemetry to dispatch

One loop, closed every few seconds.

Weather, market, and SCADA telemetry flow in; a dispatch and control actions flow out; measured grid response flows back to sharpen the forecast. The whole loop is auditable.

Now piloting with operators

Push your grid further into renewables.

HELIOS is in pilots with grid operators, utilities, and VPPs. If you are balancing a high-renewables system, we will run the engine against your network and show you the headroom.

Request a demo → Read the technical brief

On-prem & control-room deployment • IEC 61850 / DNP3 • digital-twin sandbox on request