Figure 1
Electrical and Computer Engineering
Team 2007
Team Members |
Faculty Advisor |
Brian Fomenko |
Shalabh Gupta Sponsor Electrical and Computer Engineering |
sponsored by
Autonomous drone technology has expanded rapidly into civilian market sectors over the last two decades despite primarily serving in military roles throughout the latter third of the twentieth century. Particularly interesting are the efforts of logistic-giants whose warehouses and service infrastructures are becoming more reliant on autonomous drone technology. For example, Amazon has been exploring fully autonomous package delivery by way of Unmanned Aerial Vehicles (UAVs) in recent years, with trial-runs being carried out in select parts of North America already. Developing these UAV Package Delivery Drones presents many engineering design challenges that necessitate a portfolio of diverse skills spanning different fields such as: control systems, electrical engineering, communication systems, computer vision, machine learning, and others. One such challenge is integrating a computer vision system with the navigational capabilities of the drone in the aim of being able to autonomously identify and navigate towards some object, such as a package marked for delivery or pick-up. In this project, we integrate an off-the-shelf computer vision system (PixyCam) with an existing quad-copter flight control platform (Pixhawk4/PX4) to demonstrate autonomous recognition of and navigation towards identified objects/packages.