Lean in the Warehouse?

If you are wondering how Lean plays in warehousing situations, rest easy – it does very well.  Years ago, as a materials manager in the automotive operations of United Technologies I had plenty of opportunities to apply Lean ideas in the warehouse operations of my own facility.  Over the last year I’ve had the privilege to be involved with helping out a bit at the distribution centers of a multi-billion dollar retailer to introduce aspects of a Lean approach in a number of locations performing traditional warehousing operations, fulfillment and reverse logistics.  The ‘Lean’ things we used in years gone by still work great today in warehousing situations – with the added benefit today of using value stream mapping tools to identify and drive improvements effectively in complex operations.  I’ll share a couple of real case studies a little later.

Let’s run down through the key ‘Lean’ tools that fit extremely well in a typical warehousing operation.  These are:

• Value Stream Mapping (VSM).
• 5S and standardized visual work.
• Team building, problem solving and error proofing.
• Kanbans and Pull systems.
• Line balancing and ‘cellular’ or ‘flow’ applications.
• Kaizen and good old-fashioned ‘waste reduction’ efforts.

What you won’t typically find is strong applications for some of the other Lean tools such as setup reduction, total productive maintenance, manufacturing cells and visual factory – for the obvious reason that we usually don’t ‘make’ anything in warehousing operations, other than a lot of paperwork and transactions to go along with the physical movement of goods.

Value stream mapping fits in extremely well no matter what it is you are doing.  Being a ‘material and information flow map’ as developed by Toyota and popularized in books like Learning to See, VSM is a good way to gain an understanding of how a warehouse operates – and clues to what should be improved.  My best tip for warehouse folks in applying VSM is to absolutely make yourself ‘be the thing’ as you create your data boxes for each step.  The ‘thing’ can be the SKU, order, picking line, tote or anything else that dictates the flow of product and/or information.  This concept is equally important for the physical goods as it is for the paper and information flow.

5S – which is a systemic approach to workplace cleanliness, orderliness, and standardized visual work arrangements may seem a little basic or unnecessary, but don’t be fooled.  It’s amazing how much waste occurs for the lack of 5S, or “having a place for everything and everything in its place.”  If you observe people trying to do different tasks and spending more than a few seconds looking for needed tools, information, supplies, fork trucks or anything else, it is a clear signal there is room to improve.

Unless you work in a lights-out, fully automated situation with an integrated ARS (automated retrieval system) it is likely you have some human beings who have to do real value-add work of some kind to keep the wheels turning.  If that’s the case, team building, problem solving and error proofing are universal requirements to perfect in your operations.  I think the reasons for having good teamwork is pretty obvious, but often see a real lack of attention to the latter two aspects.  If people are involved, there are sure to be plenty of vexing recurring problems such as put-away errors, counting and recording errors, accidents, spillage, damage, lost time and confusion on what to do with exceptions.  If this sounds anything like your situation, dust off your problem solving and error proofing text book and get to work!

Applying Kanbans and Pull systems in a warehousing situation may seem a bit of an oxymoron – why wouldn’t you be applying this automatically?  If you have tried it in your warehousing operations you may already know that these are not silver-bullet tools.  A robust Pull system requires a lot of work to maintain (remember, forecasts and volumes are not static) and sometimes may not be the best tool to manage certain commodities.  That said, if you are not using this technique everywhere possible you are leaving a lot of money and wasted effort on the table.  Properly implemented Kanbans/Pull systems effectively:

• Promote excellent house-keeping (5s) and the ‘Visual Work Place’.
• Improve inventory turns.
• Reduce the frequency of stock-outs and over-stocks.
• Prevent lost hours due to repackaging or breaking lots.
• Reduce expediting and premium inbound/outbound logistics costs.

If any of these are significant concerns in your warehousing situation it may well be worth the time and effort to re-examine the application of Pull systems or spend some serious investigation time finding out why your Pull systems are not giving you these kinds of results.

Line balancing and ‘cellular’ or ‘flow’ applications may sound like a weird fit in your warehousing situation, but there are many cases where these tools can have an excellent positive impact on results.  This is especially true if there are a series of things that need to happen in your warehouse – such as breaking down master packs into a retail store or distribution fulfillment requirements and having to perform re-labeling and export documentation prior to shipping.  I’ll cover how this works in a warehousing operation in a case study a little later in this article.

Finally, Kaizen and good old-fashioned ‘waste reduction’ efforts always have applications if you have more than a couple of people and processes that need to interact regularly.   Let’s look at a real situation I recently studied at the retailer I mentioned earlier.

Case Study – Applying VSM and Kaizen and Good Old-fashioned Waste Reduction in a Packing Operation

I was asked to visit a troublesome area at this company’s showcase high-volume light-pick warehousing and fulfillment operation.  Some 30,000 totes with product in them are being picked every day that flow through a conveyor system to a central pack area where the totes are auto-sorted by the computer into roller-conveyor lanes to consolidate retail store or consumer customer orders into corrugated containers.  What was reported was a significant issue with overtime and bottlenecks getting things out the door – and an urgent desire to improve productivity in packing.

The packing area has about 40 people working to pack product on 40 consolidation lanes.  We began with making observations of what was happening in the process using a standard work observation form.  In about 90 minutes and watching several different packing operators we made several observations:

• A lot of hand-writing information on the outside of boxes that did not appear to add any real value.
• Operators struggling to force boxes (cutting and tearing) into different configurations to keep the packed product snugly contained.
• Walking 30,000 totes a day back and forth from the work table to place them on a return belt.
• No standardization to the process of performing the scan-transfer-label-boxing process – there was more than a 50% variation in transfer time, based on comparing the results from operator to operator.
• Poor ergonomics and a lot of reaching and straining to see into boxes and totes.
• Other wasteful activity driven by the current process.

We went back to the office and constructed a value stream map to detail out the work cycle for the major process steps.  Some quality brainstorming and problem solving provided these improvement ideas:

• Eliminate the practice of writing order numbers and operator ID numbers on the boxes by putting the order number slips from the totes into the boxes when packing (instead of removing and discarding them).  It turns out the writing on the boxes is a ‘just in case’ we forget to label them.  After asking a few questions we learned that one or two times a month a box might get to shipping without a label.  After asking a few more ‘whys’ it became evident that we were spending over 200 hours a month trying to prevent a problem that only takes an hour to correct for the one or two occurrences of missed labels.  Further, a simple error proofing auto scan for the proper barcode label that would stop the process can easily be implemented and paid for in a few months . . . if this measure is required to convince management to eliminate this practice.
• Redesign the workstation to eliminate wasted steps moving empty totes and getting boxes and labels.  This requires a 5S presentation technique for the boxes and supplies for making boxes and shortening the workstation’s footprint to cut down on steps.
• Standardize the methods to ‘one best way’ to process totes (with the least wasted motion that is the easiest and safest).  With a 50% variation observed, it is clear that overall productivity will improve if we ‘all do it the same way, every time’.
• Eliminate the hand-scanners with permanently affixed scanners that read the tote automatically when it is positioned to be unloaded through adding some technology and workstation re-design.  I felt sorry for the poor operators – they are picking up a bar-code reader hand-gun and putting it down 30,000 times a day to read a barcode on the totes that is always exactly in the same location!
• The current process requires operators to push the finished boxes of product down roller conveyors to shipping and then carrying (30,000 a day) totes to a separate return belt.  Solution: Add an inexpensive piece of automation downstream on the roller conveyors to push the totes off to a return belt, and send the totes down the same line with the boxes.  Presto!  120,000 paces of wasted time every day are eliminated!
• Correct the poor ergonomic flow of product from totes to boxes.  Change the workstation design so that totes are tipped-up toward the packer so that they can simply ‘dump’ the product into the boxes, or at worst, be able to use both hands to move the product into a box.  This will collapse the ‘transfer from tote to box time’ by 50% to 75%, improving productivity dramatically.
• Automate the process of taping shut the boxes – this will need to be studied to verify an acceptable return on investment.
• Re-design the boxes to:

1. Include ‘inner dividers’ built into the boxes (rather than making and inserting them manually).
2. Include perforated or skip-slotted corners to eliminate the need to use a knife to make them shorter.
3. Pre-print targets for applying labels in the correct place every time on the outside of the boxes.
4. Possibly allow boxes to ‘self seal’ when set up.

The upshot?  After recording the changes to the current state value stream map it became apparent that a productivity improvement approaching 50% is possible if all the recommendations can be successfully adopted.

I feel strongly that the key to long term improvement in an area like this requires a commitment to empowering the work force.  In this case I am recommending the formation of teams of five to six people who ‘own’ eight pack stations as a ‘Kaizen Circle’ team that is charged with on-going process improvement.  Teaching and supporting theses teams to continually perfect their standard visual work and eliminate the errors is critically important.

A Pulling Module – the Power of Creating Flow Through Line Balancing

At a parts center, we spent some time being ‘the thing’ (in this case a pick ticket) for pulling product from warehouse pick faces directly to corrugated boxes for customers or stores.

The process works something like this:

• Pick tickets are processed in batches at night, based on point of purchase sales and customer orders over the internet.
• Pick tickets are then auto-sorted by the computer system by SKU and module location.
• Expeditors sort these by estimated box size and attempt to get enough pushed through in batches to keep the first pull stations busy.
• The U shaped conveyor runs about 150 feet down and back in front of pick faces.
• The boxes with pick tickets paper-clipped in place are manually pushed from zone to zone.
• Observation: feast and famine at any given pick face location at any given time.
• Observation: at the end of each day we ‘flood’ people from other pull modules that finish early to ‘push-out’ the busiest modules product during the last hour of every day.
• Observation: in all, about 17 people are assigned for the day to handle this module’s work load.
• Reported: vexing and on-going quality problems due to human error (wrong SKU, wrong qty, missing SKU, etc.).
• Sounds like your operation?

After spending some time making time observations of the ‘pick face to box’ cycle to pack the SKUs, we created a current state value stream map of the process.  By using the observed ‘value-add cycle time’ to calculate the daily lines volume into a full-time equivalent estimate we made an interesting discovery: the manpower assigned is double that required to do the ‘value-add’ work.

What?  Yes, it’s true – due to a massive macro in-balance of work between the zones in the module about 50% of everyone’s time is being wasted ‘waiting and walking’.  Now, don’t get me wrong – the people were doing their level best to make it work and literally breaking a sweat to keep things going.  People are seldom the problem; processes are!

After asking a few more questions and doing some brain-storming we discovered a surprisingly simple solution: Establishing and controlling the flow of work at Takt time.  In case you are not familiar with it, Takt time is simply a calculated ‘drum beat’ of time derived by dividing the available work time by the amount of customer demand.  For example, if you have 8 hours of work time at 3,600 seconds an hour, your available time is 8 X 3,600 = 28,800 seconds.  Now we simply divide this by the daily demand, say 2,000 pick slips.  The pick slip Takt time is 28,800 / 2,000 = 14.4 seconds.

By establishing a ‘push cycle’ of five boxes with pick slips every 72 seconds (14.4 seconds Takt per box/pick slip X 5 boxes per group), we found that we could fill the system with five boxes in front of each pick zone with a comfortable space between each group.  At 72 second intervals, all the boxes index forward on the conveyor in sets of five forward to the next pull zone and on out to shipping.  By working with the IT system to mix the daily pull slips to give reasonable balance of SKUs to pick zones in each set of five boxes it is possible to ‘keep everybody busy on value-add work’ and dramatically reduce the ‘walking and waiting’ time.

Because a 50% productivity ‘opportunity’ is now easy to visualize, the Pull module team was quick to agree that a short-term 25% improvement was realistic – and they are on their way . . .