MASLAB
MIT Mobile Autonomous Systems Lab (6.146)
🤡
Members: Timber, Camacho, Nate, Tyler
January 4th, 2022
- first lecture, talked about a lot of stuff, didnt understand most of it, it be like that sometimes
January 10th, 2022
- we showed up for lecture but there was none.
- decided that our nuc would be vertical in order to save room for our ball collection and redistribution mechanisms.
January 12th, 2022
- 4:30AM: watched a very long video about open CV and YOLOv4, just have to learn how to train it to see Balls and Walls 😏😏
- made a program that takes webcam input and uses YOLO to identify things.
- 1:30PM: Sat in lecture hall for an hour after lecture and got some 'za as a team.
- Lab: robot finally moved, however, it ran away then proceeded to only moved in circles, so we gave him a dunce cap for being a silly goose.
- after adding some wheel marker to track their speed, we deduced that one motor is wired backwards and they are backwards, i.e. left is right and vice versa.
- it does actually move and turn 4 times in a row, however, the turn is not 90º.
January 13th, 2022
- R 🤡B 🤡T has finally got his top plate! we also got a battery, finally feels like our team is not behind, lol.
- Glued the kill-switch to the side of the bot so it is easily accessible.
- From my very poor understanding heres how our code will work:
- camera (node) > camera data (topic) > object detection (node)
- object detection (node) > ball position (topic) > driving code (node)
- encoder (node) > encoder data (topic) > driving code (node)
- driving code (node) > motor commands (topic) > motors (node)
January 24th, 2022
- After many days, finally taught my computer how to see balls, however, i now realized, after talking to a class mate, that color segmentation would have been quicker, easier, and likely more accurate.
January 25th, 2022
- Working on implementing color segmentation.
January 26th, 2022
- Got our **** together, everyone went to lab for the first time since the first week, showed up at 10 am and worked until 5
- Designed a system to pick up balls by using two foam rollers with rubber bands for traction
- Redesigned and recut the side bracket for the foam rollers at least 6 times
January 27th, 2022
- grind continues
- 10am - 2am, sigma grindset dont stop
- rollers work, thats insane, first thing that actually works
- camera still doesnt work, gotta balance stuff out
- motor broken
- motor fixed
- motor broken
- different motor broken
- Timber learned how to juggle
https://vimeo.com/user164930961/review/671002385/1780cd7d6b
- 11:51PM at simmons now
January 28th, 2022,
- our roller doesnt work, sigma grindset doesnt stop.
- we are meticulpously checking our code to make sure ball detecion and drive ode worjk together to ollect balls, but our ball detection isnt working exactly as expected.