PeRL STUDIES AUTONOMOUS NAVIGATION & MAPPING FOR MOBILE ROBOTS IN A PRIORI UNKNOWN ENVIRONMENTS.
PeRL studies autonomous navigation and mapping for mobile robots in a priori unknown environments.
Naval Engineering Education Center (NEEC)
Autonomous systems are playing an ever growing, and more crucial role in the Navy active fleet. As such, the U.S. Navy has a need to educate and hire a cross-disciplinary civilian workforce who are trained in the science and technology behind the next-generation of autonomous systems. This NEEC project will crucially help to fill that role by educating and training an undergraduate and graduate student cohort in the cross-disciplinary system issues in developing robust autonomy for unmanned vehicle systems.
NEEC Project Objective
The center of focus of this project will be a multi agent unmanned vehicle system testbed that will allow the students to learn, train, and perform experiments with real unmanned vehicle systems. Proposed science and technology areas that students will work on are:
- Autonomous mapping and navigationCooperative control and searchComputer vision, Lidar and multi-sensor fusion for robust intelligenceKalman filtering, particle filtering, and Bayesian estimation methodsRobust autonomySystem integrationEmbedded system programmingSensing and perceptionControl and path-planning
NEEC Project Testbed
The testbed is comprised of one Segway RMP 200 autonomous ground vehicle, one AscTech Pelican autonomous quadrotor, and several Parrot AR.Drone quadrotors.
The rationale for choosing a dual ground/aerial platform is that:
- These vehicles exhibit all of the challenges of a real autonomous unmanned system including: embedded system programming, sensing and perception, control and path-planning.
- Having a mixed asset ground-aerial vehicle platform exposes the students to significantly different vehicle dynamics, and also allows for training and learning in cooperative heterogeneous autonomous vehicle systems.
- Ground and aerial quadrotor platforms allow for easy on-campus experimental use.
Watch the video below to see the team's progress on autonomously landing the quadrotor on a moving target. The UAV detects the target with its onboard cameras before extracting relative pose information and performing the landing maneuver.