The use of Unmanned Aereal Vehicles (UAV – commonly drones) for critical infrastructures inspections for potential damages is becoming increasingly popular, in particular in areas where the access by humans is dangerous or impossible and the examination process could become lengthy or inaccurate.
Notwithstanding the availability of UAV technologies able to satisfy and overcome several hurdles of manned inspections, long flight permitting time and high vehicle and service costs are often discouraging the end-user to adopt these means.
In addition, from one side unmanned vehicles reduces the risks related to the presence of humans inspecting the infrastructure, but on the other they could introduce: additional risks, such as air risks, impacting the surrounding population; potential increase of service costs; social concerns due to the operation of a novel technology, often associated to military actions.
The paper will dig in the questions above and suggest a solution that employs small drones embedded with RGB (Red Green Bleu) cameras combined with an AI algorithm able to elaborate the information registered by these cameras and provide useful outcomes for the assessment of the potentially damaged structure. Among these, the infrastructure analyst will be provided with recomposed images for visual inspection, smart 3D reconstruction of significant affected parts, pattern recognition and location of identified anomalies, and an approximate evaluation of some representative damage indicators, that often require knowledge of wavelengths in the invisible emissivity spectrum. Disposing of a qualitative assessment of the damages and knowing the location on the structure will facilitate the selection of more precise techniques for further investigations and/or the definition of the most appropriate remediation and corrective measures.
Small drones have several advantages. Indeed, they are less invasive, quite silent and light, thus reducing the air risks at a minimum and normally require cheap CAPEX investments. In addition, applicable rules and regulations often allow exceptions for these types of UAV. Known disadvantages of operating light drones, such as the influence of weather conditions, in particular wind and rain, can be successfully compensated by well-defined missions, appropriate flight planning and supervision by experienced pilots. The paper will discuss these issues in detail, providing evidence through a structured strengths and weaknesses analysis.