Bounty Hunting Robot

Mechatronics

Fall 2021

Project Outline

An obstacle course has been built for an autonomous robot to navigate. There are four distinct obstacles in the course. Each obstacle has unique challenges associated with it and at the ends has a wooden block bounty that must either be knocked off a pedestal, or brought back to the start for extra points. If the robot does not return to the start it is considered a failed attempt.

At the start of each attempt, the robot begins in the starter zone (bottom left) and waits for a start signal. An external arduino uses a color sensor to scan a puck which determines which obstacle to tackle, it sends this data to the robot as a start signal. Once this signal is sent, no more communication to the robot occurs.

The Hub

The robot must rotate the hub until it senses the magnets then drive onto the hub and rotate it until it reads the color of the desired obstacle

The Canyon

The robot must use line following to navigate the winding turns. It then reads the frequency of a strobe light to determine which bounty is real.

The Cave

The robot must use distance sensors to follow the side walls up to the cave. It then reads a magnet to find and then knock over the beast and open the cave. The bounty is within the cave.

The Wall

The robot must lift the wall to the top where it will stay, then reach the bounty behind.

Our Solution

Design Features

Actuators

  • Two DC motors with encoders to drive the wheels

  • Two high power servos to control the arm

  • A standard servo to drive the gripper

  • A micro servo to lower the hub rotator wheel

  • A micro DC motor to drive the hub rotator wheel

Sensors

  • Self built color sensor

  • Hall Effect Sensor to detect embedded magnets

  • 2 Reflectance Arrays for line following

  • Distance sensor for wall following

Controllers and Power

  • 2 Arduino Megas

  • 2 XBEE wireless modules

  • VNH5019 Motor Driver

  • 9v Rechargeable Lithium Battery

Early testing of the reach over approach

Results

Our robot did not place well in the competition, but we were given the "Most Interesting Approach" award for our idea to have our arm reach over the wall to grab the bounty rather than lifting the wall.

My Contributions

For the first half of the semester my main task was to design the chassis, and then to design, fabricate, and program the arm. Due to one of our members dropping I ended up doing the entirety of the programming and a large part of the wiring.


The arm and chassis were both completely assembled with laser cut acrylic, 3D printed parts, and fasteners

Early testing of the arm using potentiometer knobs to drive the servos

To design the arm I did a lot of research on how to best design the linkage so that the weight on the arm was minimized and the gripper was oriented properly. This 2-DOF linkage design allowed for the servos to be mounted at the base and transfer the power through the linkages. It also had the added benefit of keeping the gripper at the same orientation for all positions. The arm was made completely of laser cut acrylic and was connected to the base and gripper through 3D printed parts. There were many iterations that focused on improving ease of assembly, work area, rigidity, and lift capacity. Simple inverse kinematics were performed to calculate the ideal angles for specific positions of the gripper.

Laser cut template made from DXF files exported from SolidWorks and assembled in Adobe Illustrator

Laser cutting first parts of early arm design