Autonomous MAV Package Delivery for Urban Last Mile Project and Artificial Intelligence for Racecar Project
Date:
Mentor: Dr. Yiqiang Han, Mechanical Engineering
Co-Author: Wenjian Hao
Students: Phillip Do, Alexander Krolicki, Jae Eui Lee, Dhruval Patel, Nathan Witebsky, Curdin Gantenbein, Meg Nuttall, Jeremiah Ballard, Jamal Nasser, Nathaniel Powers, Steven Sheets, Duncan Nicholson, Tony Wang
2 hour interactive poster session Program
Abstract:
Electric-powered Vertical Take-Off and Landing (VTOL) technologies are around the corner and ready to be applied to commercial applications, such as unmanned delivery, emergency response, survey and patrolling, etc. Autonomous control of an Unmanned Air Vehicle for a package delivery mission is specifically considered in this project. Students in the first project will showcase their customized drone build and also the computer simulation environment for vision-based navigation model training. The objective of this project is to design, simulate and test a UAV to perform package delivery mission solely based on vision navigation and autonomous control techniques. This presentation will emphasize the simulation technique developed during this study. A “hardware-in-the-loop” testing approach was adopted. The control board and radio controller were connected to a virtual testing environment to interact with different test conditions. Dynamic response feedback can be used for testing different autonomous control algorithms. The hands-on programming and testing practices can also be applied to many other interdisciplinary projects such as self-driving car researches. The second project involves a technical display of a fully-autonomous 18th scale racecar finishing racing on a track. Guests can also try to control the racecar using a joystick and compete with the trained AI model.