Application of Autonomous Robotic Vehicles for Wireless Charging of Power Wheelchairs

The objective of this work is to provide a baseline for a novel indoor robotic application, which is an assistive robot able to autonomously recharge an electric wheelchair using WPT (Wireless Power Transfer). As such, I had to devise a plan from scratch and provide a first generation prototype to demonstrate that this application is possible following strict constraints. In particular, cost and most importantly size must be carefully considered, as the robot will have to fit under a wheelchair. This project was realized at Freedm Lab in Raleigh, at North Carolina State University. As a power electronics specialized research facility, other colleagues developed the WPT system, while I focused on path planning, communication and programming of the robot. Create 3, from iRobot company, was used for quick prototyping. A RaspberryPi 4 microcontroller was employed to control this UGV (Unmanned Ground Vehicle), running ROS 2. To test this operation a mock up wheelchair was designed on Solidworks and built, considering average wheelchair dimensions. This wheelchair model was made recognizable by the robot by mounting two IR LEDs on its back side, flashing at a particular rate defined by an Arduino Mega. Bluetooth beacons were employed in order to make path planning more sophisticated when further away, which made estimation of the distance possible. Possible future research could focus on improving the overall robot path planning performance by redesigning the current hardware, as, due to the lack of extra space, no cameras or LIDAR could be employed to create a map of the indoor environment. Another possible approach could be increasing the number of beacons.