Team Twelve/Journal
From Maslab 2007
| Maslab teams |
|---|
| Team 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 14 · 15 · 16 · 17 · 18 |
| Team Twelve’s Journal |
Contents |
Day One 1/8/07
We completed the tasks required for day one like assembling the peg bot and soldering together the orcpad. This didn't take very long and we didn't encounter any problems which was very fortunate.
Day Two 1/9/07
Today we got the camera and IR sensors hooked up. We ran into some problems with the IR sensors and our wireless usb device, but eventually these were fixed. We also agreed on a basic strategy and design plan. We hope to get most of our odometry hardware and software working tomorrow (we'll be in lab late).
Day Three 1/10/07
We got all of our sensors working (aside from the safety sensors likes the bump). We also completed the final checkpoint goals (although our ball detection isn't ideal).
Day Five 1/12/07
Today, Rian continued work on the control system and Rurik worked on modelling the different parts of the robot in SolidWorks and combining them into an overall assembly.
Day ???? 1/15/07
Today the main focus was on perfecting Washington's movement. Rian worked a lot on getting our PID controller working, and Mike worked on creating the framework for a visibility graph. The graph will keep track of Washington's position and features that are visible from that position.
Day ???? 1/16/07
The plastic for the chassis was mostly finished today. There were some issues with cutting the plastic in the lab shop, but Rurik figured that stuff out. Mike continued to work on the visibility graph
Day ???? 1/17/07
Rian and Alex wrote code to access the PC speaker. They created a basic song file format that we can use to compose music for the robot to play. The robot chassis is mostly finished today. The ball collection mechanism remains to be completed.
1/22/07
The robot was significantly finished today. We now have a ball collection mechanism and all the other physical things we'll need. Here is a picture of us and our new friend!
1/24/07
We've got the non-pegbot robot collecting balls now. Here is a picture of it in action:
1/26/07
Today was a less stressful day. Now that our robot is finally finished or approaching its final state we have been testing all of the code we wrote and correcting every tiny bug, although we have restructured our code extensively. We originally planned on having a heavily multi-threaded control program but we realized that synchronizing access to the OrcBoard was a large toll on the speed of the program. Now our program runs entirely within one thread (with very pleasing results).
We have ball aligning working pretty well and expect to complete goal aligning within the next couple days, so we have a couple more days for final testing.
1/31/07
Our robot has been assembled in its final form (with washington facemask and assembly). We have been testing and expect to test and fiddle with the robot's internals until competition day.
2/1/07
We returned our robot today for impounding. Here are the specs:
Software:
- Four state finite state machine. States: wander, scan, go to goal, and get ball.
- We have two level abstraction on top of the Orc class: OrcControl (which has methods to get data for our specific sensors) and RobotControl (which has high level methods to allow the robot to do high level things, like going to a specific "odometric" pose and going forward according to an error function. RobotStateMachine uses RobotControl.
- Although we have sensors for odometry and we drop nodes for mapping, we don't make use of it. The robot mostly randomly wanders and whenever it sees a goal and thinks it has a high probability of having balls it attempts to deposit them.
Hardware:
- Gyroscope sensor
- 3 Front positioned IR sensors
- 2 motors
- ball cage with front opening.
- camera
