Solar Data and the Promise of Solar-Powered UAVs
SolarAnywhere has long been known as a data service for utility-scale and distributed solar PV. But its high-resolution irradiance datasets are now driving a surprising new application: solar-assisted unmanned aerial vehicles (UAVs). At the Warsaw University of Technology, researcher Piotr Lichota is leveraging SolarAnywhere's historical satellite-derived irradiance data to determine the best times to launch solar UAVs and how long they can realistically remain airborne. The work could accelerate the development of autonomous environmental monitoring, precision agriculture, and search-and-rescue operations—missions where long-endurance flight is critical.
The Weather Challenge for Solar UAVs
Solar-powered UAVs are engineered to recharge their batteries with onboard photovoltaic panels. In theory, a clear-sky scenario describes a straightforward relationship between solar irradiance, panel output, and flight duration. In practice, however, weather is rarely cooperative. Dynamic low-altitude cloud cover, especially in regions like Central Europe, can drastically reduce the solar energy available to the aircraft, limiting mission capabilities.
For mission planners, the inability to predict irradiance under realistic skies is a major barrier. Most prior models rely on theoretical clear-sky assumptions or climatological averages, which fail to capture the day-to-day variability that actually determines whether a UAV can complete its mission. Lichota recognized that something more robust was needed: a simulation framework that replicates real-world atmospheric volatility to assess the probability of mission success.
Building a Stochastic Simulation on Real Irradiance Data
From Clear-Sky Profiles to Cloud-Cover Models
The core of the framework is the SolarAnywhere dataset, which spans more than 25 years of satellite-derived irradiance observations. Using Maximum Likelihood Estimation (MLE), Lichota identified clear-sky Direct Normal Irradiance (DNI) and Diffuse Horizontal Irradiance (DHI) profiles that match historical observations with significantly lower error than standard references, long-term averages, or typical meteorological year formulations.
But clear-sky profiles alone aren't sufficient in weather-variable regions. The high-fidelity datasets were used to construct a cloud cover model. By simulating the stochastic behavior of clouds, the framework moves beyond deterministic assumptions and produces probability distributions for available solar energy at any given time and location.
Modeling Mission Outcome Probabilities
With this stochastic framework in place, Lichota quantified the probability of achieving a target flight duration for a given launch time, location, and flight configuration. Instead of a single theoretical maximum, the framework outputs curves that show the likelihood of different outcomes under realistic conditions. These curves represent a fundamental shift from simple performance estimates to risk-aware mission planning.
Key Findings: Midday Launch Wins
One of the most telling results from the framework is the impact of launch time on mission success. As illustrated in the study, a midday launch offers meaningfully better odds of a longer mission than an early-morning launch—even for the same UAV on the same day. This kind of insight cannot be obtained from clear-sky assumptions, where morning and midday irradiance differences are predictable and often small.
The reason is intuitive: cloud cover patterns in Central Europe have a diurnal component. Morning fog and low clouds often dissipate by midday, allowing more irradiance to reach the aircraft. The model captures these dynamics, giving operators a data-driven basis for choosing launch windows.
Designing for Real-World Conditions
Beyond scheduling, the framework enables better design decisions. Lichota demonstrated how modifications in solar-to-wing area ratios, battery capacities, or the selection of takeoff hour directly impact mission feasibility. Instead of optimizing for an unrealistic theoretical maximum flight duration, engineers can use the probability curves to evaluate trade-offs. For example, increasing battery capacity might improve endurance under cloudy conditions but adds weight; the model shows whether that trade-off is worth the risk.
This approach facilitates the transition towards risk-oriented design for green technology applications. Rather than assuming a best-case scenario, designers can specify reliability thresholds—such as a 90% probability of achieving a certain flight duration—and work backward to define parameters that meet that threshold.
Implications for Green Technology and Beyond
Solar UAVs are just one example of how advanced solar resource data can support emerging green technologies. The same methodology could be applied to solar-powered vehicles, marine autonomous systems, or even stationary energy storage operations that depend on solar forecasting. As more sectors adopt renewable energy, the need for probabilistic, real-world performance modeling grows.
For the UAV industry, the implications are clear. Autonomous environmental monitoring missions in Central Europe and similar climates can be planned with a quantified level of confidence. Precision agriculture operations can schedule flights when the probability of adequate solar energy is highest. Search-and-rescue teams can decide whether a solar UAV is likely to stay aloft long enough to cover a target area.
Conclusion
The work at the Warsaw University of Technology highlights the value of long-term, satellite-derived irradiance data for applications beyond traditional solar PV. SolarAnywhere's dataset provides the foundation for realistic, stochastic modeling of solar UAV operations. By replacing clear-sky assumptions with probability-based frameworks, researchers and operators can make smarter decisions, increase mission success rates, and accelerate the adoption of sustainable aviation.
As the technology matures, we can expect solar UAVs to become more capable and dependable, but only if we plan for the weather as it is, not as we wish it to be. Data-driven approaches like Lichota's are the key to unlocking that future.
This article is based on reporting by CleanTechnica. Read the original article.
Originally published on cleantechnica.com
