Members: Jared Shepherd and Justin Boyce
Project Description: Local company AMBOTS has asked us to help them with improving their accuracy when 3D printing. AMBOTS performs “swarm” 3D printing where multiple 3D printing robots work in tandom to complete larger projects. The problem they are facing is there is a lot of room for error when 3d printing and currently they are using a manual solution for this. They are currently inputing readings from where the printhead is on a grid and putting that into an excel spreadsheet that compares the position where the robot thinks it is to a measured true positon. While this does currently work, it is time consuming and is succeptible to human error. They have asked us to use computer vision to improve the accuracy of their printing.
Solution: To complete what has been asked of us, we are using OpenCV, a computer vision software, to detect where the print head ACTUALLY is relative to where the robot thinks it is. To do this, the camera, attached near the print head of the robot, will find a marker placed at each of the four corners of the print surface, and triangulate where it is located within the printing surface. Then, using this information and the location that the robot thought it was at, we will put these data points into a program that will mitigate the difference and tell the robot where the print head is actually at from then on. The goal being for the team at AMBOTS to just press “go” on the software and for the process to be automated.
Task List:
Report:
Final Presentation:
Documentation: