ThingsLog VPP forecasts tomorrow's prices, builds the schedule and dispatches your plants to it — charging in the cheapest hours, selling into the evening peak, and never exporting into a negative price. Start read-only on the plants you already own; turn on control when you are ready.
ThingsLog VPP makes it real — a vertical, fully European platform, engineered for security and reliability on our own hardware, firmware and IIoT stack.
Every number below comes out of the same price store the platform plans with — the 2026 day-ahead auction, not a brochure figure.
A rough number from real Bulgarian day-ahead prices — not a promise. Move the sliders, then ask us what it would cost to run.
Assumes one cycle a day on the average 2026 day-ahead spread between the two cheapest and two most expensive hours in Bulgaria (€146 / MWh, from actual auction results), 88% round-trip efficiency, 5% downtime and a 60% capture ratio because nobody forecasts perfectly. Excludes grid fees, degradation and any balancing revenue. We do not guarantee revenue — we guarantee the schedule.
The VPP runs on ThingsLog's own IIoT platform and our 4G / NB-IoT IoT data loggers with live telemetry — the hardware that measures, watches and acts at every plant. We aim to be the technical operations & data provider behind your energy trader: you keep the market relationship, we run the fleet and feed it clean, real-time data — and, where you want it, dispatch the plants to plan. A fully European solution, built end-to-end on our own vertical stack.
ThingsLog IoT data loggers stream live telemetry from every site — 4G / NB-IoT, edge-buffered, honest 15-minute data even when a vendor cloud blinks.
The same loggers reach down to the PCS and battery over Modbus — curtailment and charge/discharge scheduling, so the plan the market was sold becomes action.
We run and watch the fleet and provide the data; your energy trader keeps the market seat. Data provider & technical operations, delivered as a service.
One vendor owns every layer — from the sensor in the field to the market interface. No foreign cloud in the control loop; your data stays in the EU. The cyan layers are ThingsLog's own.
Solar and storage arrive in every brand, every protocol, scattered across sites — while the market rewards a single, dispatchable, well-forecast schedule. Bridging that gap is the whole job.
Huawei, SolarEdge, Sungrow, SMA, Deye… each with its own cloud, API and data model. Getting one honest, real-time view of the fleet is the first wall teams hit.
To earn, you must commit a schedule against tomorrow's prices and then hold to it. Manual spreadsheets don't scale past a handful of plants — and every deviation costs imbalance.
Most vendor clouds are read-only. Actually curtailing or shifting a battery to follow the plan needs a control path down to the hardware — which few platforms own.
Reconciling plan vs. actual at the 15-minute interval, per plant and per VPP, against DSO metering — that's where margins are made or lost, and where most tools go quiet.
Each is a strategy in the engine, not a slide. Turn one on, and the day-ahead plan, the live dispatch and the settlement all follow it.
The forecaster ranks tomorrow's hours; the battery charges in the cheapest solar hours and discharges into the evening peak — placed by price, not by clock.
Below your price floor the plant stops exporting: PV is held at a curtailment floor rather than driven to zero, and surplus goes into storage if there is headroom.
Time-windowed export caps and remote curtailment through our own controllers, with closed-loop verification that the setpoint actually landed.
The day-ahead price forecast on its own — three methods, replayed and scored daily so you can see how good it was, not take our word for it.
Every VPP is its own tenant: its own plants, users, market zone and platform credentials — enforced in the engine, not just in the UI.
The national grid at 5-minute resolution — generation mix, storage, cross-border flows — beside intraday and day-ahead prices, so the plan is made with the system in view.
A demo runs on live market data and a real reference plant. Bring one site and we will show you the schedule it would have followed yesterday.
From ingest to market submission to settlement — multi-tenant from the ground up, so every VPP, trader and plant owner sees exactly their own slice.
Each VPP is its own isolated tenant — its own company, users and roles (owner · operator · technical), the same isolation model as ThingsLog itself.
Plans every plant against ENTSO-E day-ahead prices and weather, then aggregates only the operating plants into the VPP's committed schedule.
Onboard a plant read-only to watch it follow the plan, or switch to full dispatch — curtailment and battery charge/discharge — per central.
Connectors for the major inverter clouds plus ThingsLog IoT data loggers and raw Modbus — add a new vendor as one adapter, no re-architecture.
A cost-controlled agent flags faulty strings and dark-at-noon dropouts, and proposes operational and trading actions — scoped per VPP and per central.
A 15-minute ledger with schedule-vs-actual, imbalance and reconciliation per plant and per VPP — down to the DSO invoice delta.
Link each central through its vendor cloud, a ThingsLog IoT data logger, or direct Modbus. Read-only onboarding needs nothing on site.
The engine builds a day-ahead schedule per plant from prices and weather, and rolls the operating plants into one VPP curve.
Plant owners review and submit their schedule to the energy trader; the VPP presents one aggregated, dispatchable commitment.
Track adherence live, dispatch or advise to hold the plan, and reconcile plan vs. actual at every 15-minute interval.
What an operator actually opens on a working day — no black box, every number traceable to a reading or a price.
| Module | What it does |
|---|---|
| Fleet | Every plant of the VPP at a glance — power, SoC, solar, alarms — and the fleet rolled up as one power plant. |
| Operations | Live dispatch state per plant, manual override with hardware-enforced limits, command verification, kill-switch. |
| Market | Day-ahead schedule versus actual, imbalance per settlement period, intraday prices, replan on published prices. |
| Forecast | Tomorrow's price curve with confidence bands, and a daily accuracy table comparing all methods against the auction. |
| Commercial | Contracts, paid hours, invoices and the settlement ledger — revenue as it was really earned. |
| Grid | The national power system live: generation mix, battery fleet, borders, prices; other zones from ENTSO-E. |
| Settings | Strategies per plant, operating limits (SoC band, curtailment floor, C-rate), vendor clouds, market zone, per-VPP credentials. |
| AI advisor | A watchdog and planner that reads the fleet through a narrow, typed API — proposes, and in advisory mode a human approves. |
| Connectors | Inverter clouds, ThingsLog loggers and direct Modbus — monitor-only onboarding first, control when you are ready. |
Everything below is ours: designed, built and hosted in Europe. Pricing follows capacity and sites — ask us for a quote.
Controllers and loggers that measure, buffer and act at the plant.
RS-485 Modbus RTU master that sits on the existing bus next to the site logger, Modbus TCP, two relays, pulse and analog inputs, LTE and Ethernet uplink, store-and-forward buffer, local web configurator.
pricing on requestSBC-class controller with Modbus TCP to PCS, BMS and inverters, BMS heartbeat keeping, closed-loop setpoint verification, LTE watchdog fail-over and SSH-tunnelled Modbus for remote plants.
pricing on requestRead: as a BESS controller and EMS →Battery-powered pulse, analog and Modbus loggers on 4G, NB-IoT or LTE-M — years on one battery, for billing meters and sub-meters where there is no power supply.
ThingsLog catalogueReach meters and sensors across a park without cabling: LoRaWAN gateway with our pulse counters, feeding the same platform.
ThingsLog catalogueThree layers per plant, plus the pieces a trading desk buys on its own.
All plants in one portal, alarms, vendor clouds and loggers — the layer everything else builds on.
pricing on requestExport windows, curtailment floors, battery setpoints — with closed-loop verification and hardware limits.
pricing on requestPrice-driven schedule, replan on published prices, quarter-hour settlement against the plan.
pricing on requestPer-VPP tenancy, your brand, capacity bands that fall with scale.
pricing on requestThe day-ahead price forecast with quantile bands and daily scoring — no assets required.
pricing on requestCapacity and power sized on your load profile and real day-ahead prices, with the payback shown for each use case.
on requestWhat your people and your systems actually touch.
Fleet, operations, market, forecast, commercial, grid and settings — for operators and for owners, scoped to what is theirs.
includedConfiguration, alarms and history for every logger and controller, with a documented REST API.
includedMobile app to configure loggers and controllers at the plant, online or offline.
includedEvery controller carries its own local web portal on the LAN — ports, Modbus register maps, timing, uplink.
includedREST API, documented MQTT device protocol, full export on request — no lock-in clause we would not sign ourselves.
includedYour plant follows a day-ahead schedule built on our price forecast: charge in the trough, sell the peak, never export into a negative price.
Multi-vendor fleets become a single dispatchable asset with one commitment, one imbalance view and one settlement.
We are the technical arms and eyes; you hold the market relationship. Your brand, your customers, our closed loop to the hardware.
Micro sites connect through their inverter cloud — no logger, no visit — and pool into a fleet you can later flex.
Vendor clouds tell you what a plant did. Following a market schedule means telling it what to do — and that needs a real path to the hardware. Because ThingsLog builds the controllers, the VPP can go beyond monitoring to genuine dispatch.
Real-time production, energy and health — read-only. Perfect for onboarding and watching adherence, but they can't curtail or move a battery.
Modbus down to the PCS and battery: curtailment and charge/discharge scheduling, so the plant actually follows the plan the market was sold.
Security here is not a page — it is the architecture, and it was tested by independent audits in 2026. What follows is what is actually in place.
Hardware, firmware, platform, hosting and data — one European vendor, EU data centres, no component we do not control.
The AI forecasts prices, prices the plan and manages the fleet: it builds the day-ahead schedule, reads the market and scores its own forecast every day. In management mode it also executes the day-ahead schedule itself — through the same limits as a manual command; in advisory mode an operator approves first.
Per-VPP isolation is checked where writes happen, not only in the interface: a fleet administrator cannot touch another VPP's plants.
Every internal API requires a token and refuses when one is missing. Secrets live encrypted; the master key is held only in memory and re-injected after every restart.
SoC bands, C-rate, PCS rating and curtailment floors are clamped in the dispatch loop itself — whatever produced the command: plan, override, or AI.
Market and grid feeds are validated for plausibility before they can steer a plan; a starved forecast announces itself instead of quietly planning the day.
We do not claim certifications we do not hold. What we claim is verifiable: ask for the audit findings and the fixes, commit by commit.
Anything with a published Modbus register map is configuration on our controller, not a development project. Inverter clouds connect with no hardware on site.
See the full list of supported devices — what we read, what we control →
Book a walkthrough with the ThingsLog team, or sign in if you already run a VPP with us.
Leave a way to reach you and we'll open the Demo VPP right away — a live, interactive fleet running on simulated plants.
