RFID Toys Homepage
Holy crap, check out the Dangerous Things store!
Forum Home Forum Home > Projects and such > Reader Projects
  New Posts New Posts RSS Feed - Position and distance detection
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Position and distance detection

 Post Reply Post Reply Page  12>
Author
Message
Pikiki View Drop Down
Newbie
Newbie
Avatar

Joined: 21 February 2006
Status: Offline
Points: 3
Post Options Post Options   Thanks (0) Thanks(0)   Quote Pikiki Quote  Post ReplyReply Direct Link To This Post Topic: Position and distance detection
    Posted: 04 March 2006 at 11:17am
Hello people,
have you seen the Honda Asimo Robot interaction with the user id cards?

Asimo can recognize the users who carry an RF identification card very
similar to an active L-TG501 tag. And also he is able to know the user
distance and relative position, so he can bow when the user approach.

Could be possible with Wavetrend active equipment to implement this
system with position and distance detection?
 

cheers!


Edited by Pikiki - 04 March 2006 at 11:20am
http://www.pikiki.com - Tecnología y Lifestyle
Back to Top
amal View Drop Down
Admin Group
Admin Group
Avatar

Joined: 22 November 2005
Location: United States
Status: Offline
Points: 2127
Post Options Post Options   Thanks (0) Thanks(0)   Quote amal Quote  Post ReplyReply Direct Link To This Post Posted: 04 March 2006 at 11:23am
Excellent question!

I know you can get signal strength measurements from Wavetrend gear... which basically equates to distance. Although it's not an actual measure of distance, it should work well enough in most instances to determine if an active tag is "near" or "far".

Once the person was close, you'd probably have to figure out another solution for position. You could use IR sensors or video analysis to figure out a more precise position.

At greater distances you could probably use a directional antenna and a rotator (like a spinning radar dish) and monitor signal strength in order to determine the approximate direction the signal was coming from. If you didn't want to rotate an antenna, you could set up 4 directional antennas and switch between them like what's shown in Chapter 9. By getting signal strength measurements from each antenna you could easily figure out the general direction the signals were coming from.

Amal ;)

www.amal.net
Back to Top
Pikiki View Drop Down
Newbie
Newbie
Avatar

Joined: 21 February 2006
Status: Offline
Points: 3
Post Options Post Options   Thanks (0) Thanks(0)   Quote Pikiki Quote  Post ReplyReply Direct Link To This Post Posted: 04 March 2006 at 11:51am

So, knowing the signal strenght could be easy to implement distance detection. Have you idea about how much precise it is?

For position detection, you can use 2 directional antenna, heading front with a 45 degrees angle between them, and a non-directional antenna for distance detection. So you can know if the user is on back or ahead and look at he.

One of my ideas is to implement this system to my homebrew robot, so I can put permanent tags around the house like waypoints (room1, room2, kitchen_fridge, garden1, garden_pool,...), so he could navigate around the house and make his way to these waypoints.

http://www.pikiki.com - Tecnología y Lifestyle
Back to Top
amal View Drop Down
Admin Group
Admin Group
Avatar

Joined: 22 November 2005
Location: United States
Status: Offline
Points: 2127
Post Options Post Options   Thanks (0) Thanks(0)   Quote amal Quote  Post ReplyReply Direct Link To This Post Posted: 04 March 2006 at 12:02pm
Hmm... well, the Wavetrend reader also lets you set a software gain feature, which lets you adjust the sensitivity of the antenna. I'm not sure if the gain feature effects the signal strength output or not... it might just be a filter mechanism for read events based on signal strength, without actually adjusting the gain.

I've not really used the RSSI (signal strength) feature of the Wavetrend reader, but I'm pretty sure it won't be precise enough to come up with an actual distance measurement. You'd be much better off using some kind of IR range-finding or ultrasonic range-finding device for distance. Those are much more precise.

Signal strength is just that... signal strength. It has nothing to do with distance. Also, because its radio energy, it can travel through walls. Combine those two issues and you've got a case where maybe the robot thinks its 20 feet away from the kitchen, but there is a wall between it and the kitchen... that's one obstacle. Now assume the wall has metal pipes in it that were somewhat blocking the tag's signal... then the robot moves a couple feet to the left and the signal strength from the tag gets much stronger... the robot thinks its much closer to the kitchen now, but in fact it has only moved a couple feet.

Using RFID for navigation would be good for very general direction finding over longer distances. It could be used to point the robot in the general direction it needs to go in. It doesn't work so well for actual range-finding (distance measurement). The robot would need to rely heavily on other sensors to navigate to these way-points.

Amal ;)

www.amal.net
Back to Top
Pikiki View Drop Down
Newbie
Newbie
Avatar

Joined: 21 February 2006
Status: Offline
Points: 3
Post Options Post Options   Thanks (0) Thanks(0)   Quote Pikiki Quote  Post ReplyReply Direct Link To This Post Posted: 04 March 2006 at 12:17pm

Yes, actually my robot have several IR pair sensors and a movable ultrasonic sensor for obstacle avoiding and distance management.

But the idea of the RFID waypoints would be to know the direction and distance of his destination. Not just the exact point but the little area around the tag (<=3 meters?)

Amal, do you thing it could be possible to implement it? I am using an old laptop aboard the robot and a BS2 as brains and controller.

http://www.pikiki.com - Tecnología y Lifestyle
Back to Top
amal View Drop Down
Admin Group
Admin Group
Avatar

Joined: 22 November 2005
Location: United States
Status: Offline
Points: 2127
Post Options Post Options   Thanks (0) Thanks(0)   Quote amal Quote  Post ReplyReply Direct Link To This Post Posted: 04 March 2006 at 12:24pm
I think with some careful software design to catch large fluctuations in RFID distance calculations, combined with data from your other sensors, you could implement it.

The software controlling the robot and analyzing sensor data and making decisions will be the key. It must be intelligent enough to overcome many different problems and properly handle unexpected sensor data and still make the correct decisions.

Keep me updated!


Edited by amal - 04 March 2006 at 12:25pm
Amal ;)

www.amal.net
Back to Top
prashantkgupta View Drop Down
Newbie
Newbie


Joined: 19 March 2006
Status: Offline
Points: 4
Post Options Post Options   Thanks (0) Thanks(0)   Quote prashantkgupta Quote  Post ReplyReply Direct Link To This Post Posted: 20 March 2006 at 9:05pm

Hi Amal,
What we are trying to do is:
1. We want to locate the VI's (visually Impaired). That we do using a voice initailization system.
2. then we ask him his destination.
3. we have a database (the floor plan). So we find the path from the source to the destination. This path is divided into sub goals. and we issue auditory cues to him which direct him to the sub goals and eventually to the final destination.
4. What we require is a system where we (i.e. at the server) can know this VI has reached a particular sub goal ( which is a static location say a door...etc). After we know this we then issue him the next Auditory cue ( using a text to speech engine) to direct him to the next sub goal (or static location) and so on and so forth.
5. The next part of the project is Obstacle Avoidance. for which we shall need distance measurement.
So can you please tell what kind of RFIDs we may use to implement this? All we require is to know that the VI has arrived at a certan point so that we can issue him the next auditory cue. ( Basically we need to catch that event---Event hadling)
Hope to hear from you soon.
Once again thanks alot.
Cheers!

Team Drishti.

Back to Top
amal View Drop Down
Admin Group
Admin Group
Avatar

Joined: 22 November 2005
Location: United States
Status: Offline
Points: 2127
Post Options Post Options   Thanks (0) Thanks(0)   Quote amal Quote  Post ReplyReply Direct Link To This Post Posted: 20 March 2006 at 9:49pm
Using RFID might be an expensive undertaking because readers that have the range you need can be expensive... and deploying them at each doorway and "sub goal" would be much too costly. However, if you're looking to deploy RFID, you should look into two options:
 
Passive RFID
Using passive RFID tags means the VI does not have to carry around a thick, bulky RFID tag with them. The tag also does not require a battery, so they have no limit on lifespan. The only problem with passive tags is limited range. Passive tags range in frequency from 125KHz up to 900MHz. Generally speaking, the higher the frequency, the longer the read range. However, the higher the frequency, the higher the cost. You could probably use 900MHz RFID technology to read a VI's tag as they came within 3-6 feet from a reader (depending on reader and antenna hardware).
 
Active RFID
The advantage to active RFID is range. You can give each VI an active RFID tag and setup active RFID readers at each sub point (or use a single reader with multiple switched antennas as shown in RFID Toys on page 236). You can attenuate the antennas to limit their effective range according to the needs of each sub goal point. The disadvantage to this system is that the tags are much more expensive and they all use battery power (giving them a limited lifespan). Even though tags generally have a 5 year lifespan, managing and replacing tags over the long term might be too much of an administrative hassle and might cost too much in the long run to be replacing tags all the time.

Hope that helps!
Amal ;)

www.amal.net
Back to Top
prashantkgupta View Drop Down
Newbie
Newbie


Joined: 19 March 2006
Status: Offline
Points: 4
Post Options Post Options   Thanks (0) Thanks(0)   Quote prashantkgupta Quote  Post ReplyReply Direct Link To This Post Posted: 21 March 2006 at 9:14pm

Dear Amal,
What we had initially thought of was that all the doorways would be having a Passive Tag. The VI would be carrying an Active Tag. A reader would be connected to the central server and depending on the size of the target building one or more antenna for reading the ID's from the active and passive tags.

I had thought that when the Vi comes near the passive tag, it would be emmitting its ID, and that should trigger the passive tag within its range. Both these ID's (Active & Passive) are caught by the antenna and read by the reader. This way we know where the VI is currently.

In case the Active tahg triggers two or more passive tags, the received Signal Strength helps in determining that to which passive tag teh Vi is close to and the distance from that tag.

My query is whether it is feasible and if yes then what hardware should we be looking for. By this scheme i figure that we end up with the cheapest network given the passive tag can be triggered by the active tag...

Cheers
Team Drishti

Back to Top
amal View Drop Down
Admin Group
Admin Group
Avatar

Joined: 22 November 2005
Location: United States
Status: Offline
Points: 2127
Post Options Post Options   Thanks (0) Thanks(0)   Quote amal Quote  Post ReplyReply Direct Link To This Post Posted: 22 March 2006 at 8:37am

Hmm... well, there are a couple problems with your plan. Active tags do not trigger passive tags. Readers are the only devices that can activate and read tags (passive or active). Tags cannot activate other tags.

The difference between an active tag and a passive tag is only that an active tag uses a battery and emits a signal on a scheduled basis. Active tags constantly transmit their ID numbers around once every 1 to 2 seconds. Passive tags, which have no batteries, must rely on energy they receive from the reader.

Readers are the only devices that can detect and receive ID information from tags.

What you've described above is a good idea, you just need to adjust it to work. I think your only real option at this point is to use active RFID. So, once you give the VI an active tag, you can choose to put readers at every doorway, which is expensive. Or, you can use one reader, and control which antenna is connected to that reader in order to determine which area/doorway the VI is approaching. If you know which antenna is active while receiving signals from the VI, then you know which doorway the VI is standing at/near. Check page 236 in the book to get a better idea of how that can be done.

Amal ;)

www.amal.net
Back to Top
 Post Reply Post Reply Page  12>
  Share Topic   

Forum Jump Forum Permissions View Drop Down



This page was generated in 0.219 seconds.