AI in the aerospace industry: insight on the latest hardware and case study on visual drone detection

Místnost E105 - 18:35

Nowadays robotic platforms can autonomously take decisions to face with unpredictable obstacles and challenges related with a chaotic environment. Recent techniques in the Artificial Intelligence (AI) yield to outstanding visual applications yet available to the big audience. Cameras can be used as stand-alone or complementary sensors to enable higher degree of autonomy. The tradeoff between dedicated hardware resources (AI accelerators) computational power and software architectures complexity demand is a current filed of research, which we walked through during the development of a real-time visual "drone detector". During the presentation, we will also share our experience with AI in aerospace industry.

Giuseppe Inchingolo

Currently working in Honeywell as a Software Engineer, in the Sensing Guidance and Navigation (SGN) business unit. MSc. Aerospace Engineering granted at the University of Naples “Federico II”, thesis on UAVs (quadcopters) visual detection and tracking. Co-author of a research paper “Airborne Visual Detection and Tracking of Cooperative UAVs Exploiting Deep Learning”. Believes that cameras could leverage the autonomy of small drones (UAVs) enabling functional industrial applications and exciting user experience, at no cost for general public privacy and safety. He is a foody and keen on swimming.

Pavel Badin

Pavel works as a Software Engineer at Advanced Technology department of Honeywell. His work is mostly focused on digital signal processing, machine learning and SW development for embedded platforms. He really enjoys being a part of a team that works with the newest technologies for the aerospace industry. Key work experience: High Altitude Ice Crystal Detection with Aircraft X-band Weather Radar and Pilot State Monitoring, AI deployment on various HW platforms. He believes that technologies will make our lives better. His hobbies: travelling, music, spending time with friends.