Research Technology's Fit For You

Sept. 1, 2003
One of the labs is run by Tompkins Associates and is dedicated to helping clients prove material handling and supply chain technologies before applying

One of the labs is run by Tompkins Associates and is dedicated to helping clients prove material handling and supply chain technologies before applying them in their facilities. The other is managed by CHEP, the pallet pool specialists, and is designed to help both CHEP and its customers understand how palletized product interacts with various forces and technolgies in the supply chain. MHM interviewed the principals of both organizations during the course of these tours. Those Q/A sessions follow. -- Tom Andel, chief editor.

Tompkins Development Lab
Tony Gerace, partner

Jim's [Tompkins] vision was to invite clients here to introduce new technology to them. Pursue whatever technical risks they saw with us collaboratively so we could help them mitigate the risk and prove technology. This includes material handling hardware and software. We have two warehouse management systems here and we're moving toward a program where we'll be able to run a working warehouse with one WMS, flip the switch to another, and be able to show a client the differences between the two. The idea is to show them in a no-risk environment what it's like to touch and feel the WMS.

MHM:So, the correct choice depends on the circumstances of the client.

Gerace: One may be more costly in a particular type of application than the other. We go through a gap analysis. By having two functional WMSs here we can show what it means to not have that function. You may have a client that's interested in Manhattan Associates and is concerned about how it would work in their environment. This is a functional warehouse and we wanted to keep the demo real.

MHM: How do the vendors feel about their products being tested outside of their environments?

Gerace: Sometimes they like to be a part of it. But when it's just a shoot-out, what's the best technology, we help the customer determine that.

MHM: Then you give the vendor feedback about why they might have lost?

Gerace: Absolutely, and I think we've helped them improve their products.

Jim Tompkins, president

It's also a cost savings for us because if you had to go out in the field and prove the product, you have people on the road, you have to run back and forth, it can be a pain. And there's a tendency to not do it as thoroughly as you would here, which allows us to provide a more reliable solution.

MHM: I imagine the Wal-Mart announcement has created a lot of buzz about compliance with their RFID requirements.

Tompkins: Because we're going to have to go out and either recommend or not recommend RFID equipment, you really need to be technically smart enough to understand its capabilities. There's a lot of hype that it will do everything for everybody, but there are some engineering steps you must take up front to make sure it works.

MHM: But it's not a choice when it comes to Wal-Mart.

Tompkins: There will be a lot of suppliers to Wal-Mart who will be compliant with Wal-Mart but won't get any benefit themselves because they haven't worked through the engineering. We intend to be way in front of that. We're not far enough in front of the marketplace to really do the value-add because at this point the economics are still questionable. We have to make sure we really help the supply chain and don't wind up adding cost to it in total.

MHM:So you're approaching the market showing how they might benefit from the RFID technology as well as Wal-Mart?

Tompkins: RFID, to pay economically, must be a supply chain solution, not a retailer solution. Yes, the retailer will get the benefit, but do they get it and wind up paying two cents more per item or does the supplier wind up losing two cents per item? We want to make sure they have the most economical application. There's a lot of hype and that's not consistent with our approach.

MHM: Are we in a better situation now with RFID than we were years ago when bar coding first came out?

Tom Singer, principal

Standards will come out but it's a matter of who will compromise where with the proprietary technology to develop to the standard. With bar codes, when compliance became an issue there were conflicting standards and various bodies addressing those. It's the same with RFID. You have the push from the Wal-Marts of the world and you have various organizations lining up and the standards situation will sort out. We're seeing that with the specification being developed to support Wal-Mart, the Electronic Product Code approach.

Jerry Ose, partner

It's a lot more organized than it was when bar codes came out. The Auto ID Center seems to be taking the lead in pushing this forward. There's a lot more momentum and a lot more buy-in. There's a coalition of industry and technology coming together. Even with all that, it has not been accepted in the world as the total standard. Europe has a different wattage they allow for, so we're still seeing splinters. The other thing that's not common at this point are all the interfaces.

Gerace: It's going to happen but it will take longer than everybody is anticipating. Bar coding was a much simpler technology. It didn't have all the interfacing issues. Those will have to be solved with RFID. There are cost factors and a whole set of changes that go all the way to the bottom of the supply chain. For it to succeed the packaging of the products will have to be modified.

MHM: So suppliers won't be ready by 2005?

Tompkins: By 2005 you'll see a lot of people taping RFID labels to their cases, just like we did with bar codes. Back then people put bar code labels on even if they weren't readable. The same will happen here. There will be people who know if they don't ship RFID compliant to Wal-Mart they'll get fined so they'll have Charley and Bob in the back taping an RFID label on that's never been read to the system and hope that it gets read when it gets to Wal-Mart so they're compliant. There won't be a benefit there and that's not the economics of RFID.

Ose: You bring bar codes into the DC and you have a line of site and unless someone screwed up the print on it, then you'll read 99.999% of the time. How much accuracy will we give up when you don't have that line of site when you can aim that laser right at the bar code? Now you have to depend on free air to transmit the data. There are a lot of things that can hinder it including water or other fluids. From the supply chain point of view, let's say it's working beautifully with my product and marketing decides we could sell more of these if we put a nice foil package on it. All of a sudden your read rate goes down. No one looked at it from a packaging perspective to see how it would affect the read rate. If you lose a tenth of a percentage point in a DC, that's big.

Singer: When Wal-Mart's announcement first came out, we thought, wow, it's down to the item level. Wal-Mart quickly clarified and progressed their way down to cases. The item level will occur when the technology evolves to make it beneficial for all parts of the supply chain.

Gerace: We're giving our clients, through this facility, operations and maintenance training. An operations manager can see how sorters work. There's real hardware and software for our help desk technicians. The systems we put in are tied in to a help desk in this facility. They have access to technology right around the corner and the people at the help desk can leave their seat and run a test on some of the technology to emulate what's happening in the client's facility. We're also tied into the client's facility so via the control software you can dial in from our help desk, take a look at what they're doing operationally and see how it looks in the demonstration lab. It also helps our technicians in the debugging process.

Paul Faber, partner

Depending on the access rights the client granted us we can dial in and look at the data directly. It saves a lot of trips.

Tompkins: In some cases we've identified the problem before they did.

Gerace: Sometimes they're looking at a symptom of the data and they're not really looking at the root cause. They might think their server is backing up but what really happened is the read rate dropped on the scanner. Once you walk them through the process you've turned them into a better maintenance manager or technician. We're trying to eliminate the debug in the field, and a really good integrator doesn't debug in front of the client. They plan the system out on paper what the system will do. You never take into consideration all of the possibilities of what will happen, but if you do it right it will come together. As a consequence you implement systems faster and for less cost. Find the problem on paper, it costs you a buck, find it in the lab, it costs $10, find it in the field it's $100.

Tompkins: Each one of us individually has taken a broader view, where in the old days Tompkins Associates would blame problems on the material handling equipment guy, well now I'm the material handling equipment guy. At the end of the day, it's us, and the sucker has got to work on the busy days as well as the slow.

Dick Rigling, principal

As the client is going through our process where we continually review how you want your system to work, we're simultaneously educating ourselves on what they expect the system to do. At the same time we're taking the client through his process and forcing him to take the time to think through how he's going to make this work. Invariably in every project the original set of requirements or the process we were satisfying for the customer ends up changing significantly simply because both they and we are more enlightened on what they're trying to do and what the ramifications are. It's not just a collections of screens and equipment, it's a highly integrated, complex system, all parts of which must work properly, including the people operating the system. One of the fallacies people have when doing a conversion from the old manual DC operation of 15 years ago to a highly automated version in the 21st century is you no longer have the ability to force things to happen by throwing more people at it. Once you get outside the system constraints, it will degrade quickly.

Jim Capece, partner

As we go through our process we will put together a detailed transition to operations plan. What are all the elements that will have to come together before you think of flipping the switch on?

MHM: What about when business changes? How can you maintain flexibility without investing a huge amount in new systems?

Tompkins: One major women's fashion retailer, for example, defined this envelope, and we pushed them and pushed them, then implemented for this envelope, but with all the hooks. Now the presentation we're making to them calls for no change in building, no alteration to the present operation and no downtime. We'll just add capacity with the existing square footage and we're ready to go. That is with an increase in 25% of the stores that they had no visibility for. The money we save them exceeds the money spent for the application.

Faber: You do your requirements, put in the system, and lo and behold the business environment changes. One key advantage we have is by controlling the software that talks to the ERP system and then down to the material handling equipment, if the business rules change, we have flexibility to accommodate that in a very rapid turnaround, cost effective fashion. Typically you would not have to do a WMS mod, we could do a change in our control system that would protect your investment in the WMS and still make the changes you need on the floor.

Gerace: People want to extend the lives of their original systems as much as possible. We're finding that with a well architected middleware solution, that will help without investing significant capital to make a significant change in operations. But a middleware system like ours will also help when expansions do need to happen. As we add interfaces via our Tompkins Control System to a sorter or all the various technologies out there, it's done and we can re-use that code for each client we move forward with. There has been a reluctance by the WMS providers to do that. We see integrators taking the steps to do that, but the WMS providers will typically treat each of those as a new interface. They're dealing with complex problems. They have a base functionality they modify to meet the client needs for the WMS. Then the interface to the sorter is affected because of the way their code is architected, so they have to rewrite the code associated with going to the very same sorter they may have gone to for some other client. Changing the base functionality changes the hooks that drive the sorters. The fact they might have used a particular sorter on one project and the same one on another project does not allow them to re-use that code for the sorter interface.

MHM: I thought WMS providers were offering products that were more configurable to user needs.

Singer: WMS vendors do a much better job than they have in the past in being able to provide without mods a solution that meets a customer's unique needs. But at the end of the day, especially when you're talking complex top tier distribution operations, this is an execution system, and to get the big bang for the buck you have to finely tune it to meet your operational needs. Base functionality is one thing, but if I'm going down to the material handling level and add another level of complexity that must be finely tuned to how the product flows through the facility, you won't get that out of the box across the board.

The CHEP Innovation Center

MHM: A major trend in business is for retailers to ask for more and more services from suppliers. Are you seeing that with your customers who supply retailers?

Brian Beattie, senior vice president, CHEP:

In the US, customers are pushing responsibility back onto the manufacturer from a logistics standpoint for things like modular unit loads, so they can ship it all the way through the supply chain in a unit load then roll it out onto the retail floor already built. Instead of building a full size pallet and it becoming a display piece they're talking about some smaller modular display pieces that can go out on the floor. These are quarter pallets, by which you can put four on one and then roll each onto the floor as a pre-made display. Your major manufacturers are building the display. You take labor out at the retail level. This kind of one-touch display is being talked about a lot.

MHM: How is your business changing as a result?

Beattie: 85 percent of our business today is pallets, 15 percent is containers. Our growth area is on the container side of the business. We're focusing on four container groupings: intermediate bulk containers. For the chemical industry, which moves catalysts, once it's spent, we pool the spent containers and take them back to the refinery to purify it and get it ready for the next batch. In automotive there are two different sizes that collapse and have a lot of dunnage. Preformed plastic holds subassemblies to minimize damage. There's a smaller container that flows from tier 2 to tier 1. Then there's a reusable plastic container for fresh produce. We have a container business now with its own leader.

MHM: IS RFID driving that?

Beattie: RFID is an added value service we can provide to help customers take costs out of the supply chain. RFID is a driver to outsource the investment in tags you may not want to do. This is a way to get better control.

MHM: Give us more examples of "better control."

Beattie: The big opportunity is flow via slipsheet or a box or some other fast moving medium. Wal-Mart receives a lot of stuff from Asia. It's all floor loaded in the container. Extremely efficient on the Asia side. Labor is next to nothing. Load it, maximize the cube in the container, but the pain is on the receiving side of the US and that is a supply chain savings opportunity. Between Wal-Mart and the club stores, there's a lot of stuff shipping floor loaded. There's also a lot shipped on slipsheets.

Our goal is to improve customer satisfaction. The best way is through asset productivity. Having the right product at the right place at the right time. We're controlling those assets and working with our customers to make sure goods flow as fast as possible through the supply chain. Delivering at the lowest cost is also important, and the foundation of that is through standardization around the globe. There used to be 40 different CHEPs. Now we'll do something once and apply it everywhere. The perfect trip takes some of the principals of Six Sigma, TQM, and focus people on solving customer problems. Our intent is to significantly improve customer satisfaction. You'll see this concept of Perfect Trip a lot more as we expand it.

We want to standardize our service offering across the globe under a positioning of hassle-free service every day. 80-90 percent of our business is a standard, quality CHEP pallet. That pallet and the service around it is somewhat standardized already but our intent is to completely standardize that part of our business and then have four or five options that customers can choose from.

MHM: I know you've done a lot of research on plastic components. Where does that stand?

Beattie: We're finding there's other technology that can do an even better job than pure plastic. We're close to finalizing a composite material that has some plastic in it. We've done a lot of trials on the lead board and blocks and we're finding you can improve the damage ratio. They've figured out how to bind HDPE along with filler fiber to provide strength and durability at lower cost. You'll see this on pallets in a few months. The cost is more than wood but our intent is to make sure we have the most stable platform so we minimize customer damage. We can absorb that cost. With the strength and durability of the composite and our new geometry, the lift truck tine will deflect off of it or just bend it without having a catastrophic failure. It's very important we maintain ownership, so we put property of CHEP on them and go to great lengths to tell the marketplace this is part of a pool.

MHM: What kinds of things are customers asking from you?

Beattie: Some customers require a maximum level of moisture in the pallet. We have heat treated pallets that meet that requirement, from 22% moisture to 28% moisture. An example is a corn flakes manufacturer or a sugar manufacturer with very thin packaging. Any moisture that would come out of a natural wood pallet and come up through a corn flakes box, there would be product damage. We provide that. On a container, a mom and pop facility might have short dock doors at their facility. They can't take a standard 105 RPCs on an inbound load. They need to go no more than six feet high.

MHM: How do you accommodate needs for RFID implementation?

Beattie: At a CHEP facility a lift truck would come up to a stack of CHEP pallets that have RFID tags in them, we would load them onto a trailer, and there will be a signal captured on a PC. Then the trailer goes to an emitter or manufacturer, and one trailer load of pallets becomes 20 trailer loads of our customer's products. Then they ship them through their supply chains to their customers. We're a good RFID partner because of the size of the network we already have in place. Instead of you doing it over and over in your supply chain, it's not a bad idea to use us. In fact, 2.5 million product loads are moving every day on CHEP pallets across the world.

MHM: Tell us about this pilot you're doing.

Beattie: We put readers in six of our locations in the state of Florida. There are 34 emitters [manufacturers] and 350 distributor locations that participated. We ship and receive from them. We have readers at our six facilities. We know pallet number 123 went out, went through the supply chain, we know who it went to, and we know where it came from. One customer put readers in their facilities because they wanted to figure their benefit. At this point we felt our customer base wasn't ready a year ago. We've been working on this for five years intensively because we think we can add value to the supply chain from the service standpoint. We tagged 250,000 pallets across the state of Florida.

MHM: What did you learn?

Beattie: We learned a lot. First, we learned how to tag a unit load. We found out a tag is not a tag, and not all tags can be read, and where you put the tag doesn't mean all tags can be read. We've looked at over 30 different tag designs. We've evaluated read, read/write, and it's taken a lot to push these suppliers at the unit load to come up with technology that really works where you can get 100% read. In the supply chain you'll need something that will work whether it has citrus, moisture or any of the barriers that could cause problems with the signal. Moisture inhibits the signal. If you have a load of soup or juice, it could be a problem. When we started the trial we knew if you had stuff loaded on here and you just put one tag on the pallet, that depending on how you were loading and unloading the pallet you may not get a good read. Our engineers decided to put one on each end so we'd always get a read. We'd have to coordinate the read so it knows that if I read both, it's still only one read. Over the last year we also figured out where we could put one.

We're in the neighborhood of 90 characters on a tag. If you're looking for robust read/write data, you need to add a battery and that adds a significant amount of cost. With the technology we're using, all the power is in the reader. The tag receives a signal from the reader that tells it to wake up and send a signal back with its information. At the PC you could correlate whatever data you want instead of doing it all on the tag.

Then, how do you make it stay on the pallet so you can re-use it? We've had to put a robust casing around it. We're going to put it on the center block, where lift truck tines have a tough time hitting. This is probably the safest spot on the pallet, and we know exactly where we can get the best read. Matrics will probably be one of our tag suppliers. They understand the unit load.

The other important area is the reader. Lift trucks go five mph, and not any reader can read something that's in motion, so speed is one issue. Let's say the lift truck driver comes over the dock plate, is on a bad angle and another vehicle is coming behind him. He has to move up a little--did he get two reads? Or if I'm staging for something else, I set it close to the dock door, and with my read range I can read it again and again. These are things you have to deal with from the reader standpoint. This is not plug and play technology. You have to understand the environment of the plant or DC. You may have automated warehouse management systems for other stuff, and you need to be in harmony not competing with that frequency. Are you screwing up every time you read to your WMS?

If you modify a lift truck there might be liability issues there. Just because you have a reader on there now you're part of a liability situation. That was more of a challenge than we planned for.

MHM: How will users benefit directly from using the tags?

Beattie: RFID tags can be read as they go through a portal, and the data are sent to a PC. Then an ASN like you use in EDI can be created at the manufacturer. The pallet RFID tag can be associated with that unit level ASN or bar code or item level.

A final issue is the enterprise system. Now that you have the ability to read the data it needs to go someplace. You need to package it in a way that your team of people can use it. You have to deploy people to take data and use it. Now that you're measuring you have to do something with the data. To match it up with your system, whether you're a manufacturer or distributor, you have to match up with legacy systems and the beauty of it is that it has the ability to be real time vs batch.

MHM: What about the price of the tags? Is it coming down?

Beattie: Everybody's talking about a five-cent tag. They've been correlating them to the case or item level. Everyone is trying to build a business case around a five cent tag. We're dealing with millions in our world, not billions in the item level world, so we can't find anyone who can validate this real low price tag. It's not there. It will take billions to approach five cents. Fortunately, we're on the supply chain side, not the item level side, and that's where the ground swell is right now, at the pallet level. For more information on Tompkins Associates, go to For more information on CHEP, go to