B2J2

A robot to autonomously collect magnetic cubes

Comment

Inspiration

For the capstone second year mechatronics design project an autonomous robot was required to be designed, built, tested and improved in six weeks. Through the six weeks many previous skills learned in school such as software development and design process skills needed to be applied. However many new skills were learned along the way that led to the completion of our robot that we named B2J2 for a star wars theme as well as the first letter of all our names, Baxter, Bridget, Jon, and James.

What it does

B2J2 was tasked with scanning a square of unknown size to detect and retrieve magnetic tesseracts in a sea of non-magnetic tesseracts. After it scans the room it must be able to pick up the magnet and return to its starting position. At the starting position there is a platform with designated drop off positions that the tesseract must be placed on. It has to complete this task three times in under four mins. An additional problem that needed to be addressed was two robots would operate inside the square at the same time and thus a collision avoidance system is required to prevent robots from interfering with each other.

That was just half of what the robot had to complete. The robot also had to switch to what was called "Mode 2", and in this mode was a complete separate list of tasks that needed to be completed. In mode 2 it had to pick up the tesseracts from the holding position and drive under a 19cm x 25 cm structure to place it on the roof on the same structure 45cm high. All of this had to be completed autonomously with no interaction from a user.

How I built it

It was built and designed in Solid works before any mechanical pieces were put together. After multiple iterations the final design was created. It was believed that the circular chassis and arm with claw design was the best suited design for completing the tasks. For the prototype the circular chassis and arm supports were laser cut from 3/16 inch acrylic, and the robot arm and claw were 3D printed. All electronics used on the prototype were Vex components with the exception of the custom arduino. The MSEduino was an arduino based micro controller that was designed in class using Eagle and then later manufactured and assembled in a previous lab during the year. On the PCB it had two microcontrollers, 4 photometers, 4 push buttons, 4 switches and 16 charlieplex lights and 2 microcontrollers with up to 32 ports combined.

How it Worked

B2J2 utilized one ultrasound sensor, 5 Hall effect sensors and the two encoders to scan and navigate the square along with finding the magnetic squares. It would drive in straight parallel lines back and forth until one of the three Hall effect sensors along the base of the chassis detected a magnetic flux. The claw then lowers to pick up the tesseract. Depending on the orientation of the robot it would use a returnHomeLeft() function or a returnHomeRight() function. Once it travelled home and faced the platform, the ultrasonic sensor would position B2J2 in the correct position to drop off the tesseract.

Challenges I ran into

One major problem faced during the testing stages of the robot was getting the robot to drive straight. To drive straight the robot used PID control using the feedback information from the encoders. This method worked well however was extremely dependent on the power source. The constants used in the code needed to be constantly updated manually to accurately drive straight. On the day of the showcase the constants were not perfect and was unable to complete all tasks. To improve the drive straight function extra ultrasonic sensors will be placed on the left and right of the chassis as an extra set of parameters to keep the robot straight.

Accomplishments that I'm proud of

All of it!!! Even though the robot never fully worked in mode 1 or mode 2 reliably, it did preform successful test runs. Before this project no one in the group had experience in building an autonomous robot and after six weeks we were able to design, construct and test a robot that could perform most of the tasks it set out to do. Many challenges arose and long hours of debugging and iterations led to the final design. I learned a lot about mechatronic design and teamwork throughout the six weeks and I am proud of what the team accomplished.

What I learned

SolidWorks Control System Mechanical Design Gear Design Sensors Version Control Ardunio Actuators

Built With

Share this project:

Updates