Periodic Reporting for period 1 - FlightAI (Flight plan artificial intelligence assistant)
Berichtszeitraum: 2021-09-01 bis 2023-05-31
The project's main goal was to address the need for expertise and the time-consuming activities for compiling a SORA report to obtain drone flight authorizations. Completing a SORA risk assessment demands extensive knowledge in Unmanned Aircraft System (UAS) operations and a significant amount of information from multiple sources, including airspace and ground area characteristics, population density, critical infrastructure presence, specifications of the used drone, intended operation details, operator capability, and personnel competence. Manual information collection can be time-consuming, often causing delays when operators need to quickly respond to client inquiries about the feasibility of specific services. This issue is particularly challenging for non-aviation professionals, who comprise the majority of drone operators.
At the beginning of the project, EuroUSC-Italia, a partner of the consortium, provided a web-based solution called SAMWISE (www.online-sora.com) to assist UAS operators in developing SORA analyses for planned drone flights. However, this solution requires expert users to manually gather and input a substantial amount of information. Additionally, there no feedback is given to suggest alternative solutions if the operator is unable to meet the identified requirements at the end of the SORA analysis.
Therefore, the project set out to achieve the following objectives: (1) Integrate an AI search algorithm from a previous EU FET OPEN project into the current SAMWISE system; the resulting product, called 'FlightAI', willl aid humans in constructing possible drone flight plans based on user-defined via points, while reducing the risks assessed with SORA. (2) Test the commercial potential of the FlightAI product based on a validation assessment involving relevant stakeholders, and gather indications from them to determine the market's viability; (3) Carry out communication and dissemination activities to reach potentially interested stakeholders and prepare an effective marketing strategy for the FlightAI product.
The system redesign aimed to overcome challenges faced by the initial solution, such as manual route optimization and time-consuming data loading. To achieve this, advanced artificial intelligence algorithms were integrated into the system for automatic route optimization based on ground data like population density and presence of critical infrastructures. A user-friendly graphical interface was implemented to simplify the desired mission description, and complex calculations and errors were eliminated based on automatic computations of SORA risks. Automated mechanisms were incorporated in the system to streamline data loading, for example on the features of the used drone, and by automatically retrieving relevant territory information from GIS sources.
The technology assessment involved testing, simulation, and user trials involving UAS and risk-assessment experts to evaluate the solution's performance, functionality, and usability. Valuable feedback was collected and analyzed to identify areas for future improvement.
Dissemination activities, including press releases, focus groups, surveys, and a Final Workshop, effectively raised awareness and engaged stakeholders. The project's impact reached a wide audience, and the Final Workshop provided valuable insights from experts on criticalities involving drone mission planning involving both technical aspects, territory features and regulatory context. The consortium partners converged to a Joint Venture Agreement laying the groundwork for future refinements and the commercialization of the product.
The project also played a pivotal role in nurturing AI talent within European higher education systems. Through strong collaboration between CNR and AI2Life, it facilitated the smooth transition of AI talents from academia to industry, empowering the FlightAI product's functionality with AI algorithms. Additionally, by bridging the research-industry gap and fostering a proactive technology-market environment, the project contributed to the global success of European industries.
QBT sagl, a technology solutions provider involved in the project, significantly improved drone route planning capabilities with advanced algorithms based on GIS data and supporting AI route optimization. The company also strengthened its collaborations with AI2Life and EuroUSC-Italia, the other two involved companies.
EuroUSC-Italia, a prominent player in unmanned aircraft systems and regulatory compliance, played a crucial role in enhancing the safety and compliance aspects of drone operations and technology validation. Their expertise in airspace regulations and risk assessment led to the development of robust methods incorporated into the FlightAI product, ensuring adherence to regulations and the highest safety standards. The project achieved significant advancements in risk mitigation, enabling safe and controlled drone operations.
Overall, the integration of SAMWISE with AI search algorithms and user-friendly interfaces greatly enhances the efficiency and capabilities of conducting risk assessments for drone operations, resulting in safer and more efficient drone flights in Europe.