Final deliverables

Final report link:
https://drive.google.com/file/d/0B7BR6As8f2U9ZFRaM182eTFsNEU/view?usp=sharing

Github link:
https://github.com/ese-519/multi-robot-area-covering

Demo videos: (also in previous posts)
Two robots
Three robots

Change Magnet Board to a Larger One

Since we have done the basic thing for the multi-robot area covering, we began to change the board to one in a larger size with the selected round, strong magnets shown as above. Then we start to change some codes and make the three robots work in this board.
It is quite good!

And we also do charging for the recharged NiMH battery.

Three robots working

This video shows our three robots working together. Robot 1 and Robot 2 are working, and Robot 3 is charging and waiting for instructions from master. When Robot 1 stops at a certain point, it sends message to master telling its current position and its final destination. Master will send corresponding information to Robot 3 and tells its start position.

The major problem is that our robots don't have sensor to detect collision now, but they have to move on the magnets to locate themselves. So we need to plan the moving route in the future so that three robots won't collide with each other along their way.

Two robots working together

This video shows how two robots are working together. At the beginning, the master node sends out instructions that robot 1 is working and robot 2 is charging. Then robot 1 moves in zigzag trying to cover the area, and robot 2 is waiting in charging area. When robot 1 is "running out of charge", it sends a message to master with its current location. When it receives master's feedback, it starts moving to the charging area. In the meanwhile, robot 2 gets master's instruction with the position to start. Robot 2 moves to the specific position.

Robot Charging Issue

As for the robot recharging issue, we have two choices for the charging of batteries. The first one is to use voltage transformer LED and diode to build a charge and discharge circuit. The second plan is to use the wireless inductor charging.  The two choices are shown as below.

But there are drawbacks for both of them. The first one's circuit is simple but it is hard for the robot to find the charging point and connect to it automatically. The second one is easy to locate but the output current is very small and therefore we need to build a circuit to enlarge DC current.
We are still working on it.

ZigZag movement

The robot is moving in zigzag along the magnets set on the board. And every time it makes a turn, it stops a bit to adjust its heading.

Straight movement

3pi robots are differential drive robots. There are some minor difference in the rotation speed of two motors. With the IMU BNO055 mounted, feedback is received and the system is closed-loop. The robot is adjusting the motor speed along the way it is moving according to the heading angle data given by the IMU. The straight line on the ground in the video is just for reference.