SPARQEE an Autonomous Robot

Constructed from laser cut hardboard to maximize trade-off between strength, weight, and ease of manufacturing

Four axles (eight wheels) allows us to cross gaps in the competition surface without needing bridges

Two links combined with a rotating base provided three degrees of freedom, maximizing our possibilities for picking up an Ewok

Efficient low-power microcontroller responsible for processing of sensor inputs, actuating motors and communicating with the Raspberry Pi or a UART connection

High-power processor capable of quickly analyzing a stream of images from the camera using Computer Vision and communicating the results to the Blue Pill over a UART connection

Motors were powered from a separate power supply than the main processors and controlled through optocoupled signals (using both homemade and IC optocouplers), allowed for complete electrical isolation, reliably eliminating power rail noise

Series of passive and active filters to isolate and detect the intensity of 1 kHz and 10 kHz infrared signals emitted by the beacon on the competition surface

IR reflectance sensors were used to measure our robot’s position relative to the tape and to detect the edges of the competition surface

OpenCV was used to filter the image and extract contours, so we were able to identify whether or not an Ewok was present in the frame of the camera, and if so, the location of the center of the Ewok

Using vector mathematics, the position of the Ewok within the image was converted to the servo motor angles required to position the claw directly above the Ewok

A PID algorithm was used to ensure our robot continuously corrected for any deviation from the tape