GE 423 Final Project

May 11, 2016

Group 4

University of Illinois

Kyle Mitchell

Jon Husak

Tad Logullo

Jason Troutner

The final contest for GE 423 was to create an autonomous robot, which while traveling to five predetermined x and y locations in a course collects and returns blue and orange golf balls. The robot must be able to avoid obstacles and detect the difference in color between the golf balls, unloading them into two pads outside the arena for points. In order to finish the course successfully, the robot would need to retrieve at least 2 golf balls, and send the color and x/y position of the golf balls to a LabVIEW application.

Figure 1: Robot car used in competition

In our final design, we agreed as a team to make LabVIEW live update with obstacles found in the course. We also decided to add diagonal driving to the Astar guidance protocol. The LADAR sensor on the robot would be used for obstacle avoidance, while the camera would be used to find the different color golf balls. Two servos were available for the design of the golf ball gripper, and thus we went for a two gate design for the gripper. The team strategy for the final contest was to split the individual components of the robot design by group member. The responsibilities of each are outlined below.

Tad Logullo - CAD and build the ball gripper used by the robot, using two servos to capture and store the golf balls

Kyle Mitchell- Implement code for obstacle avoidance and path planning, as well as the ball pick up and drop off routines

Jon Husak- Refine the LabVIEW application, including adding golf ball location updates, obstacle displays, and course map

Jason Troutner- Create correction for “fish-eye” effect of lens, refine color recognition for the golf ball location, compute ball locations based on vision data

Figure 2: CAD of Ball Gripper

Figure 3: Built Ball Gripper

Our final robot was able to fulfill the required goals of retrieving at least two golf balls and displaying their initial location on the LabVIEW application. Pictures of our LabVIEW interface and a link to the competition video are shown below, as well as the zip file to all the code used in the project.

Figure 4: LabVIEW Display with Obstacles in Green and Ball Locations in Orange