Project Mad Hatter - Reflections

Greatest achievement

Being the first team to be graded and to rock the session!

What worked

Alice was remarkably robust!
Labeling EVERYTHING.
Krazy glue.
Helping out other teams and looking to them for help.
Clear communication about how code modules that were being written by different people would interact with each other.
Drive Trains:
- Use a belt drive with adjustable slots in the base (for tensioning) because it allows for flexibility
- Flatten the wheel rod (using the grinding wheel in the PRL) so that the set screw doesn't come loose
- Use rigid couplers between the motor and the gear shaft. This eliminates any slack in your system (so that PID can do its job correctly)
- Build your final base from the beginning. Even though it may take a little more time, it's worth it because you won't need to rebuild later
Have robot rules: always put the robot on a stand such that the wheels are free to rotate when you put it on the table (this may seem silly, but it ensures that your bot will never needlessly rolls onto the ground)
Take lots of pictures as you go! This will make documentation easier in the end.
Maintain a diligent GoogleDoc with daily updates on ToDo's (and a table of contents at the top for organization)

What we would like to improve

Allow for our our robot to recover when another robots hit it, especially as it is turning!
Make a more whimsical solution for the puncher arm. After designing so many iterations, it was sad to let go of them and go for the fastest and easiest to implement.
Splitting up the work a bit more evenly and rotating roles (so that people who are good at mechanical could do nitty gritty code and those who are good at code could try their hand at mechanical design and lasercamm work). However, given the time constraints, this would have been very difficult to implement.
Use a few more new circuits. Beacon sensing would have been challenging and interesting, but didn't seem wise to take on in the time alloted.

Gems of wisdom

Have fun!
Keep simplifying your design!
Finish early so that you can test on the game board.
Communicate with your team.
Make your project look good.
Use the laser camm.
Design for adaptability and flexibility. Changes will need to occur, so allow for flexibility in your mechanical, software, and electrical from the very start.
Sometimes ordering expensive components that you KNOW will work, as opposed to a cheaper option that only has a good chance of working can save you 2-3 days of project time. 2-3 days is invaluable. Almost always opt for the expensive option that comes with less risk.
Lay out team goals clearly from the beginning and don’t hesitate to remind your teammates about what you agreed on.
Try to accomplish at least one task every day.
Set your own deadlines that are stricter than the project check-off requirement deadlines and stick to them, even if it means staying up late early on in the project.
There usually isn’t time for prototyping/many reiterations, so when building boards or mechanical systems take the time to do it right (i.e. solder the board!), and be done with it.
Try think of creative debugging indicators (LEDs work great) that allow you to debug electronics outside of software.
ALWAYS make sure your set screws are tight.


	GEMS OF WISDOM Greatest achievement Being the first team to be graded and to rock the session! What worked Alice was remarkably robust! Labeling EVERYTHING. Krazy glue. Helping out other teams and looking to them for help. Clear communication about how code modules that were being written by different people would interact with each other. Drive Trains: Use a belt drive with adjustable slots in the base (for tensioning) because it allows for flexibility Flatten the wheel rod (using the grinding wheel in the PRL) so that the set screw doesn't come loose Use rigid couplers between the motor and the gear shaft. This eliminates any slack in your system (so that PID can do its job correctly) Build your final base from the beginning. Even though it may take a little more time, it's worth it because you won't need to rebuild later Have robot rules: always put the robot on a stand such that the wheels are free to rotate when you put it on the table (this may seem silly, but it ensures that your bot will never needlessly rolls onto the ground) Take lots of pictures as you go! This will make documentation easier in the end. Maintain a diligent GoogleDoc with daily updates on ToDo's (and a table of contents at the top for organization) What we would like to improve Allow for our our robot to recover when another robots hit it, especially as it is turning! Make a more whimsical solution for the puncher arm. After designing so many iterations, it was sad to let go of them and go for the fastest and easiest to implement. Splitting up the work a bit more evenly and rotating roles (so that people who are good at mechanical could do nitty gritty code and those who are good at code could try their hand at mechanical design and lasercamm work). However, given the time constraints, this would have been very difficult to implement. Use a few more new circuits. Beacon sensing would have been challenging and interesting, but didn't seem wise to take on in the time alloted. Gems of wisdom Have fun! Keep simplifying your design! Finish early so that you can test on the game board. Communicate with your team. Make your project look good. Use the laser camm. Design for adaptability and flexibility. Changes will need to occur, so allow for flexibility in your mechanical, software, and electrical from the very start. Sometimes ordering expensive components that you KNOW will work, as opposed to a cheaper option that only has a good chance of working can save you 2-3 days of project time. 2-3 days is invaluable. Almost always opt for the expensive option that comes with less risk. Lay out team goals clearly from the beginning and don’t hesitate to remind your teammates about what you agreed on. Try to accomplish at least one task every day. Set your own deadlines that are stricter than the project check-off requirement deadlines and stick to them, even if it means staying up late early on in the project. There usually isn’t time for prototyping/many reiterations, so when building boards or mechanical systems take the time to do it right (i.e. solder the board!), and be done with it. Try think of creative debugging indicators (LEDs work great) that allow you to debug electronics outside of software. ALWAYS make sure your set screws are tight.