0
\$\begingroup\$

I am carrying out a project in which a trolley or carrier follows a person and follows him continuously wherever he moves inside a shopping complex or at airport.

I am confused whether to use RFID or some other technology which provides me identification of person and measuring of distance as I have mentioned. If I am able to get all requirements using RFID alone or by alternative method welcome. I am a completely newbie to RF area.


\$\endgroup\$
  • \$\begingroup\$ With usual RFID you can only detect if the person is within range or not. It's not easy to actually determine where the person is. You'd need to do triangulation using RFID to actually detect location and from what I see in your description, that probably won't be suitable. \$\endgroup\$ – AndrejaKo Feb 15 '13 at 18:19
  • \$\begingroup\$ RFID coupled with a PID for the motion control using the RSSI, would be a good approach. \$\endgroup\$ – Anindo Ghosh Feb 15 '13 at 18:20
  • \$\begingroup\$ @AndrejaKo I'd think a trio of directional arrays would work, wouldn't it? \$\endgroup\$ – Anindo Ghosh Feb 15 '13 at 18:20
  • \$\begingroup\$ @Anindo Ghosh From what I can see, OP wants to have the setup on a mobile platform. Wouldn't you need space between the arrays to work well? I fear that sufficient space would not be available on such a platform. It would probably be enough to determine which antenna is the closest, but to actually get position? I could see it working as a part of a larger system that wouldn't rely only on RFID, but then RFID solution loses its elegance. \$\endgroup\$ – AndrejaKo Feb 15 '13 at 18:26
  • \$\begingroup\$ Actually no, the system wouldn't need to know position, just direction - the distance:speed relationship would be provided by RSSI + PID. The same PID would correct for direction drift, so even direction precision is non-critical with suitable PID damping. \$\endgroup\$ – Anindo Ghosh Feb 15 '13 at 18:29
2
\$\begingroup\$

I haven't done this, so this is just thinking out loud, but I'd try coded IR first. The person would wear a belt or something with IR LEDs are various angles.

The cart would have two sets of receivers on it arranged so that each receiver can determine a rough idea of the direction it received the coded IR transmission from. This could be accomplished, for example, with a few IR detectors in each receiver, with each detector having a narrow reception angle. The two receivers would be spaced far enough apart so that when the cart got close enough to the person the parallax would be great enough to reliably discern it from inifinite distance.

If the same IR message with the right code isn't received from both receivers at the same time, or there is significant intensity difference between the two, the reception is ignored like it never happened. The transmitter would send often enough so that missing a few individual receptions causes no real harm.

\$\endgroup\$
2
\$\begingroup\$

I think it's important to clarify if the area is a shopping market or a much bigger airport hall. Not only for the area but also for the lower ceiling and presence of metallic objects (shelves) that could interfere.

One basic technique is to fix a RF emitter to the person/ object (it can emit a short burst every second or so). If you place at least three receivers and can measure the signal strenght with enough accuracy, some basic mathematics will tell you where the source is. Again, this technique works the best in an open space with little electrical perturbations. For different targets, you need different frequencies, and your receiver will also need to be multifrequency.

If you cannot afford a transmitter to be attached to the target for the need of batteries and need a passive transponder, like an RFID tag, then I think (I'm not sure) that the range is very short (a few meters) so you can easily need thousands of them attached to the floor, isn't it a little bit crazy?

\$\endgroup\$
0
\$\begingroup\$

Well I'm not sure whether its possible even to measure distance along with direction using RFID tags after surveying for some time,like someone said is it cost effective to use bluetooth or any other technology such as triangulation,etc

Range required is enough if it is between :2 to 3m but i need identification and detection of point of turn, i.e:When a person turns in between two shelves the trolley should locate the position of turning and turn precisely without hitting obstacles(even though obstacles can be managed).

\$\endgroup\$
0
\$\begingroup\$

I have a friend who tried to make a shopping device for blind people, where the blind person could walk through a shop and be informed about all the products on the shelves by RFID tagging. Unfortunately, however, the RFID equipment only had a range of 20-30 cm, even though the equipment manufacturer claimed it would work for 2-3 meters.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.