Navigation and dynamic obstacle avoidance for UAVs in cluttered indoor GPS-denied environments

With the evolution of unmanned aerial vehicles (UAVs) in recent years, more and more researchers are setting their sights on the application research of indoor environment. Indoor applications include industrial facility inspection, warehouse inventory management, health sector, search and rescue, among others. However, the use of UAVs in these applications requires continuous high-accuracy positioning and pose information, and consequently an efficient obstacle avoidance algorithm. The current implementation uses Visual Inertial Odometry (VIO) to compute pose information, rapidly exploring Random Trees (RRTs) for path planning, and the PX4 stack for navigation. The goal of the project is to design an obstacle avoidance system which can avoid both static and dynamic (slow and fast moving) objects while executing optimal path planning strategy.

Faculty Supervisor:

Igor Gilitschenski

Student:

Partner:

SOTI Inc

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services

University:

University of Toronto

Program: