Tuning laundry to the right radio frequency

It started with the Managing Director of White Rose Laundries, Ejaz Osmani, asking me to look into RFID technology to further reduce his already low loss rate of customer flatwork.

At first the request puzzled me as we had already upgraded to Barcode Technology and I couldn’t see the benefit of using a different asset tracking system. Was barcoding failing us in some way? Did we incorrectly implement the technology? An investigation was required.

Tagging

My first stop was our sorting and tagging area. After observing our processes, I discovered that the barcode tags were technically doing their job and registering the items correctly into our ePOS system, SPOT. The first issue I noticed was the time it took to find the barcode labels on a large bed sheet or duvet cover. The operators had been trained to place the barcode label in a suitable corner of the linen. As we had decided on a small form factor Heat Seal Label (HSL) of 7mm x 25mm, sometimes the labels themselves weren’t being spotted during sorting, resulting in the operator applying an unnecessary second label. I must stress this happened very rarely.

This in itself wasn’t a real issue as SPOT registered the item as a new item. The problem arose when we scanned the item after it had been folded and was ready for packing. If the operator at the packing station scanned the first barcode label, then we couldn’t figure out which lot the item belonged to.  As a workaround, the packing operator started looking at the customer name, which we also had on the HSL.

Over time, when the packing station operators were under time constraints, they bypassed reading the barcodes and would just look at the customer’s names and pack the lots accordingly; we were moving backwards.

Wash, Press & Fold

The next area I looked at was the wash and pressing processes.  I had heard horror stories from other Laundries of HSL’s being unreadable after an 85 degree wash cycle as well as coming off when being put through their ironer. Luckily we had worked very closely with ThermoTex, to choose the correct label specification and Thermal Transfer Printers. We had perfectly readable labels throughout the wash/press process.

Moving onto folding, I noticed that the staff were making an effort to expose the barcode label so that packing could read it. Though I applaud the team handover courtesy, it did mean that one corner of the folded linen was slightly askew. We learnt a long time ago that when it comes to retail flatwork; presentation is everything.

Coming back to the packing area, I was relieved to see the staff were scanning the barcodes and then “tidying” the folding. They did have to look for the labels on a few of the larger linen pieces.

Overall we were doing a pretty good job but as I had been given the assignment, I had to learn what benefits Radio Frequency Identification (RFID) could give us.

Introduction to RFID

Working closely with ThermoTex, I learnt that everything starts with what type of transponders you use. The transponders themselves range from 7mm to 16mm diameter. They are attached to linen either by being placed in a pouch and then, either sewn in place or heat sealed much like a HSL.

The way they work is to lie dormant waiting for an antenna to send out a signal asking them if they are “there”. When they receive the signal, the transponder wakes up and sends a return signal saying, “Yes, I’m here”. The return signal from the transponder is an alphanumeric digit set usually 15 characters long, but this can vary. Some transponders also have the ability to be encoded with a small amount of user defined data.

The oldest technology, Low Frequency (LF) was discounted due to very limited performance and controllability (mis-scanning). Current technology is either High Frequency (HF) or Ultra High Frequency (UHF). LF and UHF are subdivided into performance categories typically grouped as F1 and F2. The main physical differences are that UHF adds antenna wires to the transponder giving it incredible read distances and better bulk scanning properties. A high quality HF transponder can be read from up to 30cm away and a high quality UHF transponder from up to 3 metres away.

Once you have chosen the appropriate transponders, you then have to decide on what type of antennae and readers you need. These range from compact combined units that mount under a packing table to tunnels mounted on conveyors or Gates/Cabinets where the trolleys are either passed through or parked allowing everything in the trolley to be scanned.

If we were doing large quantities of similar flatwork, such as those from a Hotels or Medical institutions, UHF wins hands down. The speed at which it can scan when presented with a trolley full of linen can be measured in seconds to a few minutes. Compare that to hand counting and it is really the only way to go.

HF can bulk scan but the design of the antennae and readers is more restrictive (limited reading distances) than UHF. If anything it seems that UHF product designers at ThermoTex spend most of their time shielding the antennae from reading transponders that are near the antenna but not to be scanned.

As this article is not an in depth analysis of the technologies, I thought the above is sufficient.

RFID in Action

Bearing in mind we wanted to apply RFID to customer’s linen, I started putting together a process which could address the bypassing of barcoding which I discovered during my investigation, in addition to seeing if we could reduce the number of steps and save some time.
We received the linen in plastic bags with individual lots of customer’s cleaning. Customers sent linen only bags as well as assorted garment bags. I needed a system that could individually scan the linen RFID as well as the barcodes on the garments. Mark Reynolds from SPOT guided me on how to set this up and it took all of five minutes.

Though UHF can be detuned to scan individually, I feared it might read items other than the one I wanted to be scanned. In the end I went for an HF system. Using a mid-range table antenna mounted under the sorting table and a separate reader mounted on the wall and connected to a SPOT terminal, we were able to pass the linen over the table area over the antenna. If it already had an RFID transponder then this entered automatically into our ePOS. If it didn’t scan we then knew to put one on.

The reason we went with a mid-range antenna was that this had better scanning ability to the full height of HF transponders (30cm). A crumpled up, soiled king size duvet cover could be taken straight from the plastic bag and passed over with very accurate read results. The construction of the antenna was very robust with metal shielding on the bottom to prevent interference from metal objects below the antenna and any possibility of mis-scanning any items not on top of the table.

At the packing station, I chose a compact combined antenna/reader with smaller height readability.  As the items were already folded and had very little height, I was able to save a little money compared to the separate mid-range antenna and separate reader.

I presented my findings and was asked to proceed. Once the product arrived, we installed it with relative ease and connected it via plug and play on our Windows based PC’s in the sorting/tagging and packing area.

At this stage, a bit of a panic set in. Early adopters of this technology have typically been laundry owned/rented or Hotel/Medical Institution linen. We were going to put it on customers linen.  Would they accept this? What if the transponder failed? How will I know which customer the item belongs to?

When we implemented barcoding I had similar concerns, which turned out to be unfounded. As for the possibility of transponders failing, we decided to put the 16 digit alphanumeric character set in both human readable and 128 barcode form, from the transponder onto a HSL.  While we’re at it, why not continue to put the customer’s name on as well.

Again working with Mark Reynolds at SPOT and ThermoTex’s excellent support team we devised a system to do this automatically.
Everything was swiftly put in place but at the last minute, our production supervisor raised an objection; the HSL was too long.  We had been using 8 digit 128 barcoding prior to this exercise. When we switch to 16 digit alphanumeric characters to match the transponder’s set, the barcode was twice as long as our old labels.

After a bit of head scratching, I remembered looking into 2D barcodes a while back.  As the scanner prices were higher than 128 scanners, we decided at the time to stick with 128. By shelling out for two new scanners, one for the sorting area and one for the packing area, I could use a 2D Datamatrix code.

The label size shrunk back to the size we used previously. In addition, the 2D scanner did not need to line up horizontally (as they do with 128) to read the code.

Customer Acceptance

One niggling concern remained on customer acceptance. We made the decision to go ahead and if any customer complained, we would remove the transponder and just use the new, 2D barcodes. For any items that may be wiped across a person’s face, such as napkins or towels, we decided also to only use barcodes. The concern being the customer experience of wiping a transponder across their mouth at the end of a meal or after shaving might not be too pleasant.

We implemented the system several weeks ago. On the customer side, I am pleased to report no one has complained yet nor have we mis-delivered a single linen item.

On the production side, no other upgrade has received such a positive response from staff.  Usually I get comments from staff such as “I liked the old way better” or “did we really need to change?” This time, I actually got compliments on how it had made their lives easier.

Reflective Summary

The sorting process when compared to barcode scanning is much faster.  A quick wipe over the table and it’s done. No longer hunting around looking for an HSL barcode label.

Same with the packing station. A quick wipe and you instantly know which cubby hole to place the item in. The best experience was the one that didn’t come up, namely, teething – there wasn’t any. Perhaps we were extremely lucky or the combined project team with ThermoTex and SPOT had the tacit knowledge to make this happen without the usual hiccups. All I know is that I am extremely grateful for this result.

We are now looking into the ePOS side of assisted assembly for linen. This will direct the operator to which cubby hole to place the linen and alert the operator when a lot is complete. With great excitement, I shared this with Ejaz Osmani. His response, “Could you look at an assisted assembly hardware system with light bars, for linen that works as well as Metal Progetti does for garments?  Well, I guess I know what my next project is.