I currently work on vision and autonomy in collaboration with the Prorok Lab at the University of Cambridge and the Toshiba Cambridge Research Lab. I completed my master's in aerospace engineering in the Space and Terrestrial Robotic Exploration lab at the University of Arizona. I previously worked on large, distributed software systems at Facebook where I focused on fault tolerance, reliability, and performance. My current research focus is on embodied visual navigation. In other words, I'm interested in the intersection of robot vision, motion, and understanding.
Media
Visual navigation to semantic targets using deep reinforcement learningGeneralized models trained on ground robots that seamlessly transfer to aerial robotsA 5-minute overview of my MS thesis, the spinning projectile extreme environment robotIceWorm climbing in Mt. St. Helens fumarolic ice cavesIceWorm testing: taking a step in the walk-in freezerA soft-landing robot I designed for exploration of Lunar lava tubes. I precompute a ballistic trajectory and rotational dynamics before launch. A precessing, spin-stabilized exhaust nozzle serves two purposes: to keep the nozzle oriented without any onboard actuators (e.g. reaction wheels), and to modulate the net impulse of solid-fuel rocket motors.SLAM in a lunar lava tubeA DIY ion thruster based on the NASA DAWN satellite
Morad, S.D., Mecca, R., Poudel, R., Liwicki, S., Cipolla, R. (2020 May). Automatic Curriculum for Collision-Free Visual Navigation. Submitted to CVPR 2020 Embodied AI Workshop
Morad, S.D., Nash, J., Higa, S., Smith, R., Parness, A., and Barnard, K. (2019 August). Improving Visual Feature Extraction in Glacial Environments. Joint publication in IEEE Robotics and Automation Letters and ICRA. Designated as a NASA New Technology (NASA NTR NPO 51401)
