Francesco graduated in Electronic Engineering at the University of Rome Tor Vergata specializing in automatic controls with a thesis on robotics and worked for 25 years in hardware and firmware design in five companies. He is a great fan of electronics and acoustics, which has led him to work a lot on analog hardware and digital signal processing.
The working group is made up of students, PhD students and other researchers, specifically of experts in the field of artificial intelligence, automatic controls and robotics.
- In the short term, use of a quadrotor based on commercial technologies: flight controller, simultaneous localization and mapping (SLAM), new anti-collision system and obstacle avoidance.
- In the long term, quadrotor+engines for fixed wing flight, new control architecture and control laws, and adaptive estimates based on immersion and invariance techniques. Use of neural networks to speed up features matching for localization on a reconstructed map.
- Neural networks to improve localization given a map (based on ORBSlam2) and from single/stereo images;
- Adaptive flight control laws and estimation based on immersion and invariance;
- An innovative hybrid structure to change flight mode from standard to fixed wing;
- Attitude estimation based on hybrid observers.