1

I am working on a project for a client. My project is about tracking assets indoors.

Example:

I have a large warehouse with several store rooms. Each room being about 50x50 ft large.

Each room contains assets belonging to the client. An asset may be anything from simple tools like hammers, saws, gloves, uniforms to laptops to forklifts. In the case of the uniforms client would like the tags sewn into the uniforms.

Each asset is in a storeroom. I want to tag each asset such that when an asset is taken out of Store Room A my software is able to detect that an asset has been taken out of store room A.

If an asset is moved to another store room by system should be able to track that the asset was moved from Store Room A to Store Room B almost immediately and send a notification to the manager on his phone.

Also if an asset is removed from the warehouse my system should be able to detect that as well and also send a notification to the manager's phone.

At any point in time my client or the manager should be able to use the software to know exactly which store room any asset is in.

From the research I have done I know this is a system of RFID tags and Readers. However after checking prices of readers I have realised cost wise it would be too expensive deploying multiple readers all over the warehouse or for each store room.

What I would like to know is:

  1. Can I have a single reader per warehouse and still be able to extend the range so it covers all store rooms?

  2. I would need a recommendation for the type of reader to use.

  3. I would like a recommendation for the type of RFID tags to use for this purpose.

  4. Are there any other equipment I would need to complete this project?

  • 2
    This is going to be very expensive. How many assets are we talking about? How much is “too expensive”? – jcaron Jan 27 at 23:25
  • 1
    @jcaron assets will be in the hundreds to thousands. One reader per room, can have as many as 50 rooms. – user3718908 Jan 28 at 11:51
  • 1
    @MawgsaysreinstateMonica No I am not, I am just not aware of any alternate solution. I like your answer and agree with it. Can you share a link with me on how to implement a BLE solution and will that mean we will no longer be using RFID tags? If yes then what will I be using to replace the RFID tags? – user3718908 Jan 28 at 11:53
  • 1
    BLE beacons and active RFID tags are functionally similar (likewise for UWB trackers). They need a battery, and they emit a signal at regular intervals that can be received by the reader. That makes them larger and more expensive than passive tags. Passive tags are much smaller (can be wafer thin, and only a few cm wide) and cheaper (well below $1 apiece), but the detection distance is much shorter (ranges from cm to a few meters at best), and handling many different items within range is problematic. – jcaron Jan 28 at 12:09
  • 1
    @MawgsaysreinstateMonica If you want to send BLE advertisements, I'm really not sure an ESP32 is the best option in terms of wake-up times and power consumption. I believe Nordic's NRF51/52 would probably be a better option, but there are a lot of other similar chips which are more geared towards very-low-power operations. Don't forget that the chip (and antenna) may be very small, but you still need a battery, and that is quickly the biggest component in the device (at the very least in terms of thickness). – jcaron Jan 28 at 14:10
2

I think this is going to be a difficult problem to solve with the given constraints.

There are two main forms of "tags": active and passive.

  • Active "tags" may be RFID/NFC (on various frequencies), BLE beacons, UWB beacons, and more. They run on battery, and transmit a signal at regular intervals, which is then picked up by a receiver.

    Those can have a pretty large range (meters, tens of meters, sometimes even more depending on the technology), but this in itself may be an issue: unless your rooms are Faraday cages, the tag may be picked up by a reader quite a distance away. So you need to resort to geolocation, either by using the strength of the signal received by multiple readers and multilateration or kNN techniques, or using AoA (Angle of Arrival) or AoD (Angle of Departure) techniques, or more likely, both.

    Given the requirement for a battery, active tags are necessarily bigger than passive tags, and more expensive. The smallest examples would be things like Tile's Sticker (⌀ 27 mm, thickness 7.3 mm) or Slim (86 mm x 54 mm x 2.4 mm). You also have cheaper (but larger) options from APR Brother: EEK-N (39 mm x 39 mm x 15 mm) or ABN02 (86 mm x 54 mm x 4 mm) or ABSensor N01 (33 mm x 33 mm x 10 mm). A more versatile option could be Espruino's Puck.js (⌀ 36 mm, thickness 12.5 mm). There are probably a lot more, those are just the first that come to my mind. You basically have the choice between flat and large (think very thick credit card) or small and thick.

    In terms of costs, I think it will be difficult to down significantly below $10 per tag. Add to that the gateways, and don't forget to factor in battery replacement, and of course the replacement of all the tags which will be lost/damaged during operations.

    An added issue is that with very large number of assets, you may have issues with collisions of the advertisements of all those assets. Not sure how well it all scales.

  • Passive "tags" do not have a battery. They are powered by the reader through induction loops. They are mostly RFID/NFC tags on various frequencies.

    Having no battery, they can be much, much smaller, i.e. wafer thin. They can also be very cheap (less than $1 for sure, most probably less than $0.20 depending on the exact type of tag and the quantities involved). They come in all sorts of shapes and forms, but it can be as simple as a small round sticker.

    However, the range for those is much smaller. The standard range is on the order of a few centimetres. Depending on the exact technology used, it may go up to a meter, but I doubt you would be able to cover a whole room with a single reader (or even several, really). This is a lot more appropriate for scanning items coming on a conveyor belt for instance rather than all items in a room. Not sure if you could even be sure to be able to scan all items going through a regular doorway.

    See here for a discussion on increasing the range for standard 13.56 MHz tags and the associated issues.

    One issue with this option is that when the reader scans for tags, all tags in range will answer at the same time. There are anti-collision mechanisms, but even if they properly work, in many cases it just means you'll only be able to read a single tag rather than any specific tag. And with hundreds of tags, all at different distances from the reader(s), this is quickly going to be quite messy. Those anti-collision protocols are usually designed for multiple tags presented together a few cm away from the reader.

    Another issue is that such tags don't like much being stuck directly on metallic surfaces. You'll need a ferrite separator or something similar for them to work.

An alternative to the whole identifiable tag could be the sort of systems used for theft prevention. I'm not familiar with the details, but my understanding is that they are a lot more crude, and basically will only detect the presence of the tag rather than the specific tag (or its position), so they're good to detect if someone takes something through a passageway, not so much what, or what assets are in a given place. But I may be wrong on that.

Let us know how things work out!

Edit

Just stumbled on this: https://www.sparkfun.com/products/14066 — according to that page, that reader can handle hundreds of passive tags (less than $0.40 apiece, very flat, but relatively large) up to 4.9 m away with the right antenna (this probably varies depending on the region, and will probably have a narrow beam at that range).

You are advised not to stand in the radiation beam of the antenna (!).

I have no experience whatsoever with those, I would be curious to see if others can comment on the actual range and performance. Also not sure if it would solve the issue of the actual location.

| improve this answer | |

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.