What is an RFID system? Four parts, and one decision.
An RFID system consists of labels on your objects, equipment that reads those labels, software that stores the read data, and a connection to the system you already work with. That is the technology, and it is much the same with every supplier. The decision that makes the difference is not in that list: where the reading moment sits in your process. That is what a rollout fails on, not the equipment.
The parts
What an RFID system is made of
In this order, because this is how they hang together: the label decides what you can read, the reading point decides when, and only after that is there anything to connect.
- Labels
- A chip with a number, on or in the object. Cardboard, plastic, metal or washable
- Readers
- A handheld you carry, or a fixed reading point at a doorway or workstation
- Software
- Where the read data lands: what was seen, where, when, and what is missing
- Connection
- To your ERP or WMS. Often only movements and deviations, not your whole stock
What is not on that list is the most important part, because it is not a product: the reading moment. That is the place in your process where someone walks past anyway, so reading can happen without an extra action.
Where it goes wrong
The technology is rarely the problem. The reading moment is.
Almost every RFID project that stalls, stalls on the same two assumptions. The first is that the label reads everywhere: metal and liquid reflect and absorb the signal, and then it works in the meeting room but not on the floor. The second is that reading happens by itself. If there is no natural moment in your process where the object passes by, you add an action instead of taking one away. Both assumptions can be tested in half a day, and that is cheaper than discovering them after the rollout.
- Metal
- Calls for an on-metal label with a spacer
- Liquid
- Absorbs the signal, so reading happens before filling
- Bulk
- Hundreds of labels at once, as long as they are not fully shielded
The honest limit
What an RFID system does and what it does not do
Wel
- Read large numbers at once, without touching or aiming at every object
- Record when something passed a point, and what did not pass it
- Search for one specific object, with a signal that grows stronger as you get closer
- Show the differences between your system stock and reality
Niet
- Give the exact position of every object continuously, the way gps does outdoors
- Read through metal or through a filled tank
- Repair a process that is not followed without RFID either
- Pay for itself with low numbers and objects of little value
That first line on the right is the expectation that fails most often. RFID gives you the last observation and targeted locating, not a moving dot on a map. For almost every warehouse problem that is enough, but it is not the same thing and it is better that you read this now than after the rollout.
What it is about
The same system, three very different questions
The parts are the same everywhere. What you solve with them differs per kind of operation, and so does where the reading moment belongs.
The rollout
How to start without a project that runs a year
The order below is the reason we do not quote a rollout before step two is done. A price without a measurement is a guess you get held to later.
01 / Pick one process
Not your whole operation, but the place where it hurts most. One site, one group of objects, one action that comes back often. Everything you learn later, you learn here for a fraction of the price.
02 / Measure whether it reads
With your own material, on your own floor. Which label works, at what distance, and how reliably at the pace your people work. One visit is enough for this, and it is the step that gets skipped most often.
03 / Work out what it returns
A baseline measurement up front and a second measurement afterwards, on the same process. That produces a go or a no-go with numbers attached, plus what a rollout will cost in work and money.
04 / Roll out in phases
Only now come fixed reading points, gates, dashboards and the connection to your ERP or WMS. In that order, because connecting a process that does not work yet only gives you faster wrong data.
The first step
What it costs to find out
Three steps with a fixed price up front. They build on each other, so you start where your question sits and not at the first one.
- Test day
- 1,500 euros. Does it read on your material, yes or no?
- RFID validation day
- 2,500 euros. How good can it get then?
- Proof of Value
- From 3,500 euros. What does it return, and what does rollout cost?
All amounts exclude VAT. The three sit side by side on the pricing page. See the pricing.
Questions that usually follow
Is RFID not just a more expensive barcode?
For one object at a time a barcode is cheaper and often better. The difference shows up with numbers and with line of sight: a barcode has to be seen and aimed at one by one, RFID reads hundreds of labels at once without touching them. If your gain is not in numbers, barcode is the honest answer and we will say so.
Do we have to replace our ERP or WMS for this?
No. STOCKflow runs alongside it and reads what happens physically. In a rollout we usually connect only the movements and the deviations, because that is where the value sits. During a test we deliberately connect nothing at all.
How many objects do you need before it pays off?
There is no lower limit that holds for everyone, because it depends on how often you search, how many people are involved and what an object is worth. A thousand cheap objects nobody looks for return less than a hundred expensive ones that go missing daily. The calculator on this site lets you do that sum with your own numbers.
Does it work on metal?
Yes, with an on-metal label that holds the chip away from the surface. Those labels are more expensive and larger than an ordinary label. Whether it works on your specific parts is exactly what a test day measures, because that depends on shape and placement.
How long does a rollout take?
A test day gives you an outcome the same day; a Proof of Value runs from baseline measurement to decision, and that lead time is agreed per assignment. A rollout after that depends on the number of objects and reading points. The order above exists so that you only have to answer that question once you know it works.