A solar installation’s annual energy potential and its output over the next hour are different questions. The first helps determine whether a project should be built. The second can influence how an existing facility schedules electricity use, storage, and purchases.
Nova-sight is exploring that shorter operational horizon. The project illustrates how observations collected for meteorology can become inputs to an energy-management service, provided they are translated into forecasts that customers can use.
From Satellite Observations to a Forecasting Service
ESA’s Nova-sight project report, updated September 7, 2026, describes a demonstration led by Belgium’s PropheSea. The proposed service combines Meteosat Third Generation imagery with local measurements and predictive models to estimate solar production.
Its intended customers include industrial energy users, energy communities, trading participants, and system integrators. Forecasts would be delivered through a software interface, with uncertainty information accompanying the estimates.
The update reports active model development and validation against photovoltaic measurements. It also records contracts with Kortrijk Business Park and AZ Groeninge. Reception equipment had been procured, although installation remained pending approval.
Those milestones indicate development activity and commercial engagement. They do not establish independently verified forecasting superiority or quantified customer savings.
Why Observation Timing Needs Careful Explanation
Satellite imaging frequency, forecast-refresh frequency, and delivery delay describe different parts of a service.
Imaging frequency concerns when the satellite collects a new observation. Forecast-refresh frequency concerns how often the software produces an estimate. Delivery delay is the time taken for an observation or prediction to reach a usable system.
EUMETSAT’s Meteosat Third Generation documentation distinguishes a ten-minute full-disc imaging service from a regional rapid-scanning service designed for 2.5-minute observations. These capabilities should not be treated as interchangeable or assumed to describe every operational data feed.
For a customer evaluating a forecasting product, the important questions are which observations it actually receives, how promptly they arrive, and what happens when the expected data are unavailable.
Better Forecasts Must Lead to Better Decisions
The business case should be tested against a customer’s decision process.
Consider an illustrative industrial site with solar panels, a battery, and flexible equipment. Advance information about reduced solar output could help the site decide when to charge or discharge storage. However, the best action would also depend on electricity prices, battery limits, production requirements, and the cost of changing a schedule.
A more accurate forecast would not automatically produce savings if the customer had no practical ability to act on it.
For that reason, evaluating Nova-sight or a comparable service should involve both forecasting performance and operational outcomes. An assessment could compare decisions made with the new service against decisions made using the customer’s existing forecast, over the same conditions.
It should also account for integration costs and the consequences of incorrect predictions. These are proposed assessment principles, rather than published Nova-sight results.
Uncertainty Is Part of the Product
A forecast is an estimate of a future condition. Presenting it as a single confident number can conceal the range of plausible outcomes.
For an operator, the consequences of overestimating production may differ from those of underestimating it. A useful service should help users understand that distinction and choose an action appropriate to their own exposure.
Forecast quality also needs a defined comparison. Performance should be assessed separately for the time horizons and operating conditions that matter to the customer. A favorable average can obscure weaker performance during the periods when decisions are most difficult.
A Developing Application of Public Space Infrastructure
Nova-sight’s broader significance is the proposed connection between weather observations and everyday energy operations. Satellite data become commercially useful when software converts them into information that fits an existing decision.
As of September 14, 2026, the project remains a demonstration undergoing validation. The next meaningful evidence would concern repeatable forecast performance, reliable delivery, and measured benefits in customer operations.
Key Takeaways
- Separate satellite observation frequency from forecast updates and delivery delay.
- Assess a forecasting service against the decisions a customer can actually change.
- Require comparisons, uncertainty information, and measured outcomes before accepting savings claims.