K-Dog

For my masters I'm trying to create a indoor position system (IPS) for the navigation of a robot.

At this point in time, i'm just researching the all of the options. Ideally I want a non-direct line of sight system.

I've breifly looked at Bluetooth, RF, and Magnetic Fields. If anyone could please either point me in the direction to research these field further or please offer some other methods for IPS that I could look into.

The rough idea, is to have a room 10m x 10m. In two corners of the room have some sort of beacon. The robot should be able to identify each beacon, and using some method work out the distance from it to each beacon, thus calculating its x,y location in the room via trig.

The problem is the beacon. We need something that can talk to the robot, with out interferance from other devices or physical objects. Secondly, the method in with the distance is calculated must try and be accurate to aleast 10cm to 20cm.

I want to advoid signal strength methods, and focus on time of arrvial methods.

Can someone please offer some advice, options or point me in the right direction.

Cheers

K

dknguyen

Well TOF+no interference from physical objects pretty much rules out everything except for radio waves doesn't it? I guess you'd better find out what type of circuits are used to measure the TOF of light. 10-20cm is so small compared to 300 000 000m/s though. It's gonna be tough to get that resolution. Might be a bit tough on the clocks too if you're thinking about local GPS.

Or perhaps a phase shift method might work. If you're lucky maybe you can use a long enough wavelength such that you can assume it's always the first wave you are looking at for your small 10m space rather than a wave more than one wavelength away from the transmitter and having to figure out which wave it is.

Or maybe you could ignore the "interference from physical objects" and use sound which is much slower and easier to work with for the distance measuring part and use radio for the communicating part. How bad does the space echo?

K-Dog

Its a standard room, so the echo is minimal. We arn't limited to rf though, http://acl.mit.edu/prigge/ this was an article i found on using magnetic waves.

rf is looking like the affordable option right now...

richardv2

Here's an article I wrote a while back. It rotates one laser on the vehicle; receives return signals from three passive reflectors and accurately computes position.

http://www.seattlerobotics.org/encod...9/lasernav.htm

mrmonteith

Look up the operating principles of a GPS. You could probably do the same thing using sound, seeing that would be easier to detect changes.

K-Dog

Hey, thanks for the advice. The laser navigation looks promising.

Cheers

K