This week we meet for the first time as a team. During this meeting we discussed our individual roles on the team and began working on brainstorming designs for the different aspects of the robot.
This week we began work on our website. For this we laid out the basic structure of the site and filled in a few sections. We also continued our design discussions and began to narrow down our ideas toward a final design. Finally, we also reviewed the game rules in depth and used that information to update our design ideas.
Jacob laying out design ideas
This week our team focused our efforts on 3D modeling our design idea and prototyping various components. Our goal with the 3D model of our system is to make sure that all the purchased and intended components can fit within the robot's size limits while achieving the functionality we require to win the competition. Below you can see a screenshot of our current progress with the model.
Current 3D model of robot (at beginning of week 3)
In addition to modeling our design in software, we have also began the process of prototyping and purchasing components for our robot. These prototypes include 3D printed parts and a wooden stand for us to mount the launching mechanism. By creating this simple prototype, we are able to test how far our proposed launching design is able to launch the ball along with other factors we need to understand to purchase the correct equipment for the final launching subsystem.
Dexter the dog inspecting the prototype
Nicole cutting out boards for the prototype
Finally, we were also able to create a working actuator prototype. This actuator will be used to adjust the angle the robot is launching.
Working actuator prototype (Credit to Jacob)
Our plan for the rest of this week and the following week is to finish our first draft of the 3D modeled design and have ordered almost every component we need for the final robot. The 3D model should be as accurate as possible but it is expected that updates and changes will be made once we start manufacturing, assembling, and testing the different components.
Working sensor-actuator combo:
Webcam detects the AR tag and motor rotates in direction that the AR tag is detected.
This past week, we worked on the CAD, acquired an Raspberry PI, developed functioning AR detection code, tested and calibrated the AR detection, created a base for the robot, and attached a motor to it. We created a working sensor-actuator combo with a camera and stepper motor. The camera was calibrated for the AR detection and the stepper motor moves/rotates towards the direction that the AR tag is detected. This is for the base of the robot rotating in the x plane aiding the catching.
Robot base with motor attached from underneath the table
Begin manufacturing the body and main components of the robot. This will involve 3D printing multiple pieces and machining by hand pieces of either wood or metal. The main goal is to complete the lazy suzan design and have it attached to the motor and have it function properly.
This week we have made significant progress on our robot. Jacob wrote the AR detection code that we will use to track where our human player is during the competition.
Jacob writing AR detection software
After verifying the code with a few basic tests, we all worked on the design and construction of the base of the robot. This involved a lot of time in the shop cutting out material, drilling holes, and assembling.
Jacob drilling mounting holes in the base
Nicole wiring the motor for testing
Once the base was assembled, we attached the AR detection unit in order to conduct more in depth testing. These tests involved using the AR detection system and a stepper motor to align the robot accurately with our human player (who will be wearing an AR tag during competition).
Basic testing setup
This video shows the webcam detecting and the motor turning in the direction that the AR tag is detected.
Finally, we also began wiring all of our motors in preparation for the next phase of the build.
Amy wiring our multiple stepper motors
Throughout the design and build process we have encountered a few challenges. Our first major challenge was designing an easy to use and build launching system. We discussed many different design options but ultimately settled on a cam driven system. We chose the cam system due primarily to cost concerns and simplicity. With the cam system, we will be able to build the majority of it by hand from extra materials we already have and will be able to 3D print any of the remaining pieces for a very low cost. It also requires less unique hardware than the other options we considered which makes the assembly and troubleshooting steps significantly easier.
Our second challenge occurred while building the base of the robot. Our large, lazy Susan style bearing that allows the robot to rotate and orient itself was binding when rotated more than 30 degrees. To fix this, we disassembled the base and found the two issues that were causing the binding. First, we had a bolt head that was sticking out too far and dragging as base turned. This was easily fixed by counter sinking the hole and making sure the bolt was secured. The next issue was a slight misalignment of the motor coupler to the top piece of the base. This was fixed by simply remounting the coupler, making sure it was as perfectly centered as we could make it.
Our plan for next week is to begin manufacturing the middle section of our robot. This includes the launching system and housing for the majority of the electrics. We anticipate that portion of the build will be challenging and time consuming due to the number of components that need to be included. We will also be working on the code for launching and a ball detection system that will automatically tell the system when the robot has successfully caught a ball.
-Finished CAD model of cam assembly
-Obtained linear rail, created and 3D printed a linear rail holder and cam components
-Constraint: Unable to create wood parts for catcher system without wood shop sticker
-Jacob obtained wood shop certification, marked on wood where to cut out
-Obtained 2x4s and 2x2s of wood for middle section of robot
-Finish manufacturing middle section of robot as pictured:
-Put together 3D printed parts and the linear rail for the cam assembly
-Cut out marked parts on wood
-Attach 2x4s, 2x2s or dowels
-Create the net for the catcher
-Create housing for electronics
Catcher assembly with cam components
Marked out spots to cut wood
Printing the final parts
IT'S BUILT!
-Put together 3D printed parts and the linear rail for the cam assembly
-Cut out marked parts on wood
-Created housing for electronics
-Put together all components to finish middle section of robot
-Launcher/catching mechanical mechanism is mostly complete (just need net)
The upper section hit the mid section, was fixed by creating a new coupler, this would have interfered with the rotation in the y-z plane. We drilled holes into the catcher/launcher basket to reduce vaccum for when the ball is launched.
-Create net for ball to fall in
-Finish wiring the robot
-Test the robot
Finalized the net construction
Completed most of the wiring for the motors and the motor controllers
Worked on cable management
Continued work on the autonomous programming
Robot assembled with net
Amy soldering wires for the motor
Test, test, test!!!
Cam launcher mechanism failed to work, need new way to launch the ball
-Decided on a different technique to launch the ball using compressed air
-Went to home depot to purchase components to assemble PVC air tank
-Ordered solenoid actuated valve
Completed PVC air tank is mounted onto bottom of platform
Tested new air tank system
Completed code to autonomously control the robot
Competed in the competition with a working robot
The competition was designed so that each team got 2 chances to show off their robot and try to score points. Our robot performed all autonomous functions well during the first round but was unable to launch the ball the appropriate distance. During our second round, the robot and human player were successfully able to play catch twice before the air tank ran out of air.
With a little more tweaking, our air launch system could have worked really well. Our main problem at the time of competition was a leak in the system that we were not able to fix in time.