Swarm robot: Experiment about IR sensor and Tasks

22 views
Skip to first unread message

Leo Yan

unread,
Aug 20, 2012, 4:10:14 AM8/20/12
to xinch...@googlegroups.com
1:Experiment: to test the distance of IR-emitting and the overlap between two IR-receivers
1.1 Test environment:
      Refer to attachments.     The conical mirror is made by tinfoil paper and use transparent plastic tube as its holder.( to make the mirror maybe a bit of complex, so we can use 8 emitters with 50 degree instead of one emitter. The cost perhaps is 0.2*8=1.6 Yuan)

1.2 Result:
1): on the condition of conical mirror on top of one emitter,there can be 360 degree infrared. Maybe we should use a transistor to drive the emitter.
2): can control the intensity/distance of emitting infrared by adjusting the duty of 38K wave. (1/3 duty has the strongest intensity. Because of using LED as the indicator of receiver,the small duty of emitting pulse can't impact on the reviver.) And the distance can be controlled from 2cm to more than 15cm.
3): two IR-receiver(viewing angle is 90degree) fixed with 80-90 degree from each other can form gap and overlap viewing area.
4): for relative location, there is a problem that the reflecting infrared from other objects will disturb locating. In order to reduce the impact, we should control the unidirectional distance of emitting infrared to 10-15cm(2-3 times as length as the robot ).
2: Tasks 
        (I think we should consider the division of tasks to accelerate the project.)
2.1 Hardware tasks
1)to design sensor part (one emitter with conical mirror and four receivers;  another:  use eight emitters as one 360 degree emitter)
2)to Make 3-5 proto-robots basing on the arduino-uno.

2.2 software tasks


Functions Tasks
Moving

  • move forward,backward

  • rotate clockwise,anticlockwise

  • small swing and return to the original state. (in order to locate)

  • move with different speed (low speed)
Near Detecting

  • To get the unique ID for robot

  • To get the chance to send near-detecting message.

  • to send and receive the infrared that includes the self ID.

  • To control the radiant intensity of emitter. Less than 5cm

  • to record who or what is nearby
Relative Locating

  • To get the chance to send locating messages.

  • to send the broadcast message of 'I am ID'.

  • to send the p2p message of relative location from the neighbouring robot to myself.

  • to record the relative direction
Message Transferring

  • to make the protocol for communication

  • To get the chance to send messages.


general
  • to make the software architecture.

  • users can define new rules in the condition-action table.

2012-08-20 07.09.03.jpg
Screenshot from 2012-08-20 14:26:25.png

萨克森之力

unread,
Aug 20, 2012, 4:48:09 AM8/20/12
to xinch...@googlegroups.com
Well done Leo. That's good news and we should try to get this running on a test platform directly on the PCB in a real environment.

~Lutz

spanner888

unread,
Aug 20, 2012, 4:53:07 AM8/20/12
to xinch...@googlegroups.com
Leo

I really like your experiment and I especially like that your are so keen and wanting to accelerate the project!

I forgot to mention this idea in the recent Swarm Robot update emails, so everyone else is probably can't see from the photo what the idea is. 

In brief, this is using an upside down cone with a IR emitter (transmit) LED underneath it to send a 360 degree signal that can be used to help other swarm robots avoid collisions and also to communicate.

Would you be able to bring your experiment on Thursday night? This will help explain your idea to the others.


The idea of using a standard Arduino (eg a Uno) is also an excellent way of speeding up the project and also would help with Lutz's goal of "design another board with the Atmega168/328 ".

I have built one of the new main motor PCBs and now an IR LED shield, and it is quick to build more motor PCBs (shields take a while - we should do a PCB - there is the start of a schematic in github -- anyone want to do this??), so I think we can do as Lutz suggested - do both the Attiny & Atmega (starting with Uno etc) versions.

What we really need is for some more members to help with coding and all of the tasks below. So everyone don't be shy, grab the code from github and/or come to the space, especially on Thursday night!


...... and if anyone also wants to help documenting and doing the competition entry - I would really appreciate that as it is taking a lot of effort!
It would be absolutely fantastic if we could at the very least have some instructions how to make and use the swarm robots in other languages (Chinese of course, but French and German and any language possible) because Africa has many languages AND because one goal of the competition is also to go worldwide with the robots!

萨克森之力

unread,
Aug 20, 2012, 5:55:48 AM8/20/12
to xinch...@googlegroups.com
Thanks Spencer for your coordinating job. I am sorry that I cannot help much now.
I will try to get myself up to speed this week when I am in Berlin and have time to sit on the computer.

~Luz
Reply all
Reply to author
Forward
0 new messages