Posts

Piwars.. THE DAY!

Image
 All excited for our first ever participation in pi-wars. William gates building was buzzing with enthusiasm early in the morning. Challenges We started with pi-noon's knock-off. We gave a good fight but got knocked out in the first round. Not an ideal start for the day.  Next was lava palava challenge. We changed our wheels to standard wheels from mecanum as planned. We had to go in blind. No test runs were allowed. We were a bit sceptical on how the Robot run would go, if the sensors would work correctly, if the light was too bright, if the black was black enough.  The bot however made a perfect run in 33 seconds. Yippee!! Eco-disaster time. We practiced well for this and were all prepared. We completed the task is about 2 minutes well within the allocated 5 minutes time. Minesweeper was a piece of cake with mecanum wheels. Next was escape route. There was a last minute hiccup with organizers announcing that the walls and obstacles were only 97mm high. We planned for or...

Testing time!

Today we tested the line tracking with a longer course that has a hump and sharper turns. We decided to switch to normal wheels for this task because the mecanum wheels struggled with sharper turns and the hump. We also found that if we made the middle colour sensor less sensitive it would work better because it only tracked a small width. We also made a course for the escape route challenge and tried to use the ultrasonic sensor. We ran into a few issues though like bad accuracy and odd distance readings. We stabilized it by taking a median of five readings at each instance. Finally for the eco-disaster challenge tests, we got some empty coke cans, filled them with stones to get a weight of 46g and we are all set. Our bot and chassis are stable (even comfortable with 3 cans at a time). Test run below:

Chassis time

Image
 Today we put our craft hat on and sat together to build a chassis for our bot. We decided to build something with thin wood and a glue gun and something that doesn't block our sensors. We then brainstormed a bit on how to create an attachment on our chassis to collect the cans in eco disaster challenge. We had 2 ideas... Prototype for attachment idea-1 looked like this.. The second idea involved lollypop sticks :) Here is a short video showcasing our attachments.   Finally it is time to paint our chassis and make it look cool. Well... it looks alright. Maybe we can better it.. that's for another day.

Black is white

Image
For Lava Palava challenge,  We got a bunch of black chart sheets and painted a white strip (1.9cm) in the middle to try out 'line following'. But the sensors on our bot just couldn't tell the difference between black and white. We got really worried that the sensors don't work. We tried adjusting the sensors sensitivity but with no luck. We then had this AHA moment that our black sheet may not be black enough for the sensors. And it indeed was the case.  Workaround... black cello tape to our rescue. Lava palava challenge, here we come!!

April fools day morning.. its Pi day for us.

Image
 We want to get the remote controller to get working today. All hands on...  Oh well.. not all of us. Siddhu has decided to design cardboard attachment for eco-disaster challenge. After a million hours... remote control works.. forward, backward, sideways, diagonal and spin!!! And our bot can climb up the ramp as well. We need to focus on precision though.

Escape route - Proof of concept

Orson put together the first version of the code for escape route. Autonomous solution using distance sensor and sideways motion using mecanum wheels. Nice home setup to test the proof of concept!

Zig zag we go!

  Today we focused on getting the mecanum wheels to function correctly. We created separate functions for motor control of each individual wheel. We then made a function that takes in a (x,y) pair and speed as input and accordingly makes the robot move forward, backwards, sideways or diagonally. In the video, we are demonstrating zig zag motion (diagonal movement).