”The Aero EDIH project provided critical insights in autonomous UAV capture in the most challenging environments. With the funding support we had the opportunity to test world class equipment and software for state-of-the-art concrete image capture. Through the project we developed several new capturing methods that will be important moving forward.”, says Ludvig Emgård, Founder & CPO – Spotscale.
Challenges
Spotscale, a scale-up company based in Linköping, develops advanced software for digital inspection of industrial structures using computer vision, AI, and 3D visualization. The company focuses on detecting and analyzing concrete damage, currently particularly for hydropower assets. Pilot customers in the hydropower sector require highly efficient and repeatable data collection methods to monitor structural integrity over time.
However, achieving automated and repeatable Unmanned Aerial Vehicle (UAV) image acquisition for large concrete structures poses significant challenges due to strict proximity requirements, high-resolution imaging standards, and the need for precise georeferencing. The project aimed to validate a workflow for UAV-based image capture on a specific dam, ensuring that the method could be replicated in future inspections for accurate year-over-year comparisons.
Solutions
Aero EDIH partnered with Gerdes Solutions AB to plan and execute UAV-based image acquisition at a hydropower dam. Preparatory work included validating flight methods, planning software, and testing two new camera sensors for usability: Evo II and Fujifilm GFX100s. Additionally, the team evaluated UgCS, a flight planning software for semi-autonomous operations to ensure repeatable image capture.
After selecting the most appropriate tools, Gerdes Solutions executed a three-day mission at the dam, including safety training and coordination with on-site personnel. The operation aimed to capture high-resolution images under challenging conditions, such as poor connectivity and complex structural geometry, while maintaining safety and compliance standards.
Results and Benefits
The field test successfully demonstrated UAV-based image acquisition for concrete inspection but revealed critical challenges that must be addressed for future scalability.
Poor GPS reception at the dam site caused positioning deviations, and the tested flight planning software lacked the detail required for close-proximity navigation. Additionally, the current image collection method did not provide sufficient georeferencing for repeatability.
To overcome these issues, the project identified several improvements: adopting flight planning software with better Ground Sample Distance (GSD) capabilities, modifying equipment with an alternative lens for the Fujifilm sensor to reduce safety distance while maintaining image quality, and implementing ground control points for accurate georeferencing. These findings highlight that off-the-shelf semi-autonomous solutions are insufficient for such demanding environments. Instead, tailored hardware and software combinations are necessary to meet the extreme requirements of automated close-range photogrammetry for hydropower infrastructure.
The project strengthened Spotscale’s ability to refine its technology roadmap and provided actionable insights for future deployments, ultimately supporting safer, more efficient, and repeatable inspections of critical energy assets.
Perceived social and economic impact
This project supports the development of advanced UAV-based inspection methods for hydropower infrastructure, enabling more efficient and accurate monitoring of concrete integrity.
By reducing reliance on manual inspections and improving repeatability, the approach can lower maintenance costs, enhance safety, and extend asset lifespan. The findings also contribute to innovation in industrial inspection, positioning Spotscale as a key player in digital transformation for critical energy sectors.
Lessons learned
The project highlighted the importance for planning and executing UAV-based concrete inspections in complex environments. It’s essential to validate flight planning tools in advance, ensuring they provide sufficient GSD and support close-range operations under challenging conditions.
Incorporating ground control points for accurate georeferencing is equally critical to achieve repeatable results over time. Hardware adaptations, such as selecting alternative lenses, can help maintain image quality reducing the safety distance to the structure, enabling safer and more precise data collection. Relying solely on standard semi-autonomous flight software is not advisable for demanding environments like hydropower dams, where structural complexity and poor GPS reception can compromise positioning accuracy. Similarly, underestimating the limitations of satellite navigation in remote or “urban canyon” settings can lead to deviations and operational inefficiencies.
“An exciting project where advanced visual AI tools together with drone technology are used to inspect critical infrastructure”, says Tomas Westlund, Project Manager, RISE Research Institutes of Sweden
