Using Building Footprints Data to Minimize Ground Risk in Drone Missions — And How it Complements Population Density Data Under SORA
Building footprints refer to the precise two-dimensional outlines or perimeters of buildings as they appear on a map when viewed from above.
When preparing for a drone mission, safety on the ground is paramount. Whether your operation involves package deliveries, infrastructure inspection, or environmental monitoring, understanding the potential risk to people and property is critical. Traditionally, planners have turned to population density data to estimate how many people might be affected by a drone’s flight path. While this information is invaluable, it only tells half the story. Building footprint map data provides a powerful layer of context that, when used alongside population density information, leads to more informed and safer mission planning. This holistic approach aligns with frameworks like the Specific Operations Risk Assessment (SORA), a structured method widely used in the drone industry to ensure regulatory compliance and enhance operational safety.
Population density data remains a cornerstone of ground risk assessment. In the context of SORA, it is used in particular for the determination of the intrinsic Ground Risk Class in the operational volume, and ground risk buffer. However, without building footprint data, population density alone may lead to broad generalizations. For example, an area with a moderate population density might include both residential blocks and empty industrial lots. Population density is usually based on census data, which is usually about residential occupancy and not industrial or commercial sites, where people work and shop during the daytime.
Without building footprints, the operator misses an opportunity to fine-tune the route. An area appearing sparsely populated could still pose a higher risk if that small population is concentrated in a few buildings directly below the flight path.
To account for the potential presence of people around building footprints, one approach is to generate reasonable buffers around the footprints. So while the population density, based on the finest available population density grids i.e. 200 x 200 square meters, 400 x 400 meters or other sizes will help in assessing the level of risk, building footprint can further be used to ensure the actual flight path avoids flying over zones where people are more likely to be present.
When determining the Final Ground Risk Class, buildings could be considered as shelters (a strategic mitigation) to reduce the GRC. The JARUS SORA 2.5 methodology does not explicitly mention “buildings” as a mitigation tool. Instead, it focuses on whether people are effectively protected from an out-of-control unmanned aircraft. SORA therefore examines whether physical barriers—such as buildings—offer real protection against a drone impact or debris. Simply pointing to a building on a map isn’t enough; you must show that the structure is sturdy enough to protect occupants. Any claim that “people in these buildings are safe” needs to be supported by evidence, which also depends on your drone’s class.
That said, maps highlighting the precise locations and footprints of buildings remain useful as a starting point for your analysis. If you can demonstrate that people inside a robust structure are shielded from potential harm, you may reduce the overall ground risk classification. Ultimately, the Competent Authority (e.g., a Civil Aviation Authority) will decide whether your justification—using buildings as shelter—meets the standard for lowering risk.
Some countries lack reliable information on population densities. According to the SORA 2.5 methodology, if designated by the authority, you may use a qualitative population density. However, defining what constitutes a "remote area" (1 small building every km^2) or a “lightly populated area” (area with small farms or residential areas with very large lots) can be challenging without access to building-footprint data and an automated process to derive a quantitative population estimate.
Fortunately, building-footprint data is now widely available and highly accurate in most parts of the world. One excellent resource is the Overture Maps Foundation. They are creating reliable, easy-to-use, and interoperable open map data by sourcing and curating high-quality, up-to-date, and comprehensive datasets from multiple providers.
When it comes to building footprints, they aggregate data from many sources such as OpenStreetMap, Microsoft Global ML Building Footprints, Google Open Buildings, and other sources based on geographies. If you plan to use these datasets, note that each data source has its own licensing requirements (e.g., CC BY 4.0, ODbL), and reviewing and complying with these license terms is important when using or redistributing the data.
With extensive contributions—from individual mappers to automated segmentation of imagery using machine learning, your area of operation is likely covered. Our experience shows that even in regions with limited road infrastructure, such as certain African countries, building footprints are often well-mapped, which help in planning operations such as medical deliveries by drones.
In conclusion, I tried to highlight some of the benefits of considering Building footprint. Building footprints data offers the missing spatial context that, when combined with population density information, forms a powerful, comprehensive strategy for minimizing ground risk in drone missions. This synergy is particularly valuable within the SORA framework, as it empowers operators to make more nuanced decisions, strengthen their safety cases for regulatory bodies, and ultimately run drone operations that are both efficient and safe.
If you have questions, insights, or experiences to share, I’d love to hear from you! Feel free to reach out at yvan@stratomaps.com or leave a comment with your feedback.