seminars
Detail
Publication date: 19 de December, 2025Developing 3D Perception for Space Robotics
From Low Earth Orbit to the moons of Saturn, space poses unique challenges to remote robotic operations. In this talk, I will share my experience working at NASA JPL, developing autonomous navigation software for Mars Helicopters and Moon rovers.
Specifically, this talk will highlight several projects where classical geometric computer vision and state estimation are used to address two core navigation capabilities (i) robot self-localization on planetary surfaces, and (ii) hazard detection and avoidance.
Additionally, I will describe how our navigation software stack was tested and integrated for a flight mission. On the other hand, this talk will also cover some perception problems where Deep Learning is making ground.
| Date | 14/01/2026 2:00 pm |
|---|---|
| Location | DI Seminars Room and Google Meet |
| Host Bio | Pedro F. Proença is a new Assistant Professor in AI at NOVA-FCT. Previously, he worked at NASA JPL from 2020 to 2024, first as Postdoctoral Fellow and then as Robotics Technologist. Later, he worked as a computer vision engineer at Tekever. He received his PhD, in 2018, from the University of Surrey and his MSc, in 2013, from ISCTE-IUL in Telecommunications and Computer Science. He is interested in autonomous systems, computer vision and robot learning. |