Pick without scanning every item separately. Check without unpacking the order.
In picking, the time does not go into the walking but into the handling around it: pick up, turn until the label shows, scan, put down. And the check afterwards often takes more time than the picking itself, because you have to open the box again to see if it is right.
We first measure the time per order line and the number of errors that only surface at the customer today.
Sound familiar?
These situations come back at almost every warehouse that ships at piece level.
- Every item has to go under the scanner before it may enter the box.
- Checking whether an order is complete costs more than picking it.
- Loading errors only come to light when the customer calls.
- Returns come back with a discrepancy nobody checks any more.
- Peaks are absorbed with temporary staff who do not know the items.
What it costs
What a picking error costs per year
Order lines per day
Error rate
%
Cost per error
euros
Working days
Per year
31,464 euros
With the share of the theoretical gain you actually achieve:
- Conservative
- 12,586 euros
- Expected
- 18,878 euros
- Potential
- 25,171 euros
Worked example based onnine hundred order lines a day with an error rate of four in a thousand, at 38 euros of handling cost per error, over 230 working days. Transport, the credit note and the conversation with the customer are in there, reputational damage is not. Not a measurement at a customer. The numbers are adjustable: put your own situation in and the outcome follows along.
How picking and checking become one
The gain is not in walking faster. It is in two actions that happen separately today becoming one.
01 / Label at inbound
The label goes on as the goods arrive, on the piece, the case or the carrier. This is the design choice that decides the rest: what is not labeled here cannot be checked further down.
02 / Pick without aiming
The picker puts it in the bin or on the roll cage. Nothing has to be turned to get at a label, because the tag is read without a line of sight.
03 / Checking as one read
At the packing station or the dock, a fixed point reads the whole order at once. What is there against what was ordered, and the difference surfaces at that moment.
What RFID does and does not do in order picking
What RFID does and does not do in order picking
Wel
- Check a whole bin or roll cage in one read
- Show missing and over-picked items at the dock
- Work without a line of sight, so without turning or opening boxes
- Run alongside your existing barcode scanners, on the same handheld
Niet
- Direct where the picker has to walk: that stays the job of your WMS
- Check items that carry no label
- Read reliably through tightly packed film, metal or liquids without adjustment
- Prevent the right item going into the wrong box, when both are labeled and read
There is a limit that often gets noticed late: RFID checks what is in the shipment, not whether it belongs to the right customer. That link stays with your WMS. What does change is that a missing or over-picked item surfaces at the dock instead of at the customer.
What changes in practice
Picking without scanning each piece
The picker walks the route and the lines tick themselves off. The scan per item disappears.The check at the dock
The whole shipment in one read against the order. Complete or not, before the door closes.The error that never reaches the customer
A wrong item stands out at the check and not at the unpacking. That is the difference between a correction and a complaint.A new picker without training time
The device points the way and says when it is right. New people during a peak are no longer a risk.
What this looks like in your operation
The same solution lands differently on a production floor than in a warehouse. Pick your own situation and you will see what changes for you, and what a Proof of Value measures there.
How we prove it: a small, paid Proof of Value
One order flow, one type of carrier. We measure the time per order line and the number of discrepancies we find at the dock, next to what slips through today.
Scope
- One order flow or one customer group
- One type of bin, box or roll cage
- Temporary labeling at inbound
- A fixed reading setup at the packing station
- Baseline measurement of time per order line and of errors that now reach the customer
Go or no-go
After the test period there is a measurement next to the baseline: how much time per order line came off and how many discrepancies we found at the dock that would otherwise have gone through. If it delivers too little, that is the outcome and it stops.
What you are probably thinking
Our picking errors are already low, there is little to gain here.
Then the question is what you pay today to keep them low. At most companies that sits in a double check: someone goes through the order a second time. That check is the cost, not the error. When a read takes it over, that time disappears while the error rate stays the same or drops.
A label on every piece is far too expensive for us.
For cheap items in large numbers that is often true. That is why we do not always label per piece but sometimes on the case or the carrier, and sometimes only on the item group where the errors sit. Where the line falls follows from your own figures, and we work those through in the Proof of Value instead of pulling out a price list now.
Does our WMS have to change for this?
No. In the Proof of Value, STOCKflow runs alongside it and compares what was read with the pick list. Only once that matches does a connection make sense. The pick sequence and the walking route stay with your WMS in any case, that is not what we are for.