Hello,
i'm looking for an idea to find out the actual position of an object on a 2x2 m large plane.
I want to install in the object a battery powered signal transmitter (IR, ultra sonic,...).
Then i want to record the movement (X/Y values) of the object.
The accuracy should be < 1 cm with 10 samples/sec.
What is the name of this ... "2D position measuring"?
Are some shields existing for that or are there any other appropriate methods?
Can this "object" rotate or just move in the X and Y axis? If you don't have to deal with rotation, you can just hack up a computer mouse and attach it to the bottom to record movement along the two axes (you'll have to start at a known position, though).
--Doug
No, I am not aware that any shield does 2D position measurement like that you described. Position measurement is not an easy job. Some special setups can get 2D position data but require some restrictions. For example, some people have used an optical mouse to track position of an object in 2D. That would require the surface to be relatively smooth and the object stick close to the surface, where the mouse camera/sensor has good focus. Other times you could try IR LED on the moving object and an overhead IR camera or a Wiimote to track it. I've also seen people doing this with robots and QR code.
So without describing your application in details, especially how this object is moving and what the constraints are on and surrounding the object, I'm afraid there is no general solution to any 2D position tracking problem.
The Pozyx system claims to achieve cm accuracy, but it is not cheap.
A mouse is not an option because you loose the position if the object is not on the surface.
The object is manually moved by older people to train them.
The rotation and the height is not used. Only the X/Y position on the ground.
A camera is not very good because a hand or the arm hides the object on top.
So i could place a lot of signal transmitters and to measure the position by triangulation?
So i could place a lot of signal transmitters and to measure the position by triangulation?
Did you look at the Pozyx link?
The Pozyx is sold out. I can get it on February 2016 (that is in 2-3 months).
At the moment i triangulate with the length of two cords. That is good for testing but a catastrophe with untrained people.
Rubbernose:
So i could place a lot of signal transmitters and to measure the position by triangulation?
It's possible to do this with ultrasound. I don't have any links to share on the topic but if you can't find any I bet I could track a few down.
DuaneDegn:
It's possible to do this with ultrasound. I don't have any links to share on the topic but if you can't find any I bet I could track a few down.
Some Korean company a while ago made a product that turned a typical screen into "touch" screen. The principle is the same as the two cord except untrasound replaces the cords. Don't know where the link ism
I saw something like that a few years ago. Basically any flat surface could be a virtual whiteboard with this sensor bar set up on the top edge.
That's exactly what i need...
If it's in your budget, take a look at this:
http://www.faller.de/App/WebObjects/XSeMIPS.woa/cms/page/pid.14.17.109/agid.1514/atid.7755/ecm.at/Car-System-Digital-Beginner’s-Set.html
More about it:
http://www.faller.de/xs_db/DOKUMENT_DB/www/allgemeines/FALLER_Car_System_2014_EN.pdf
Shpaget:
If it's in your budget, take a look at this:
http://www.faller.de/App/WebObjects/XSeMIPS.woa/cms/page/pid.14.17.109/agid.1514/atid.7755/ecm.at/Car-System-Digital-Beginner’s-Set.html
What? That system sure doesn't look like it would be appropriate for this application.
Another way, cheap, is to support the surface with force gauges at the corners and use weight distribution to accurately calculate where the object is. Will this be one object on a board or will there be more objects? Think about standing on a Nintendo Wii balance board. Same concept.
If the object is a distinct colour and you can put a camera above the work surface, a Pixy will do quite well for position sensing.
A camera is not very good because a hand or the arm hides the object on top.
So i could place a lot of signal transmitters and to measure the position by triangulation?
IMHO, this leaves out video, IR, Ultrasonic, weight , anything except RF. X-ray is not am option.
Something like http://coolarduino.blogspot.ca/2015/09/radio-direction-finder-arduino-based.html
But a hand or arm doesn't hide the object permanently, does it? It should be pretty simple to allow for short periods of blocking the camera's view.
Are the positions constrained? Can you plug the object into a grid of sockets on the board? Or does it have to slide analog-style?
A graphics tablet, like a CAD input tablet can do this but finding one that large will be difficult. I believe they have a grid of wires inside the tablet that are energised in a pattern (sweep across, sweep down) the puck or pen has a coil in it that senses the current in the nearest wire so it is pretty easy to work out where it is based on the timing.
A large touchscreen can be built with conductive paper, but that senses all touches and not a specific object. And it won't work at all if someone leans their elbow on it.
Ah, ok. That could be a solution. But the "operator" could be a problem...
DuaneDegn:
What? That system sure doesn't look like it would be appropriate for this application.
Why? That system is specifically designed to provide positioning in 3D.
Shpaget:
Why? That system is specifically designed to provide positioning in 3D.
My apologies Shpaget. I should have investigated the product better before posting my assumptions. I ignorantly thought the "Car System" was for tracking full size cars.
After finding the English version of the site, I learned these are for model cars.
Again, I apologize for making such a quick uneducated assumption.