In one week, in three different countries, I encountered three examples of actions that were sold as improvements but just months later are now recognized as being part of the problem. All of them saved money compared to what had been the current state. All had reduced waste and were considered application of lean manufacturing. All involved capital investment and approval by the customer. All were seen as a best practice at the time of implementation.
All of them now prevent further improvement and need to be ripped out.
The first one involved installing a conveyor in the packing department to eliminate the need to transport cartons manually to the next station. It saved a tremendous amount of wasted motion and resulted in a 2% headcount reduction of the entire facility. The second was a mezzanine that improved flow and speed for the picking of small parts. This reduced walking and saved a significant amount of area that was then used to increase storage capacity in the distribution center. The third was an integrated system to weigh, wrap, label and move pallets of material from packing to the dock area. Automating this was a big improvement from doing it all manually.
All of these projects had reduced costs, increased capacity and increased speed. All solved a problem for a department or function, and all of them solved the wrong problem. Now we had to explain to the customer why we want to take out the automation they paid for just a few months before, and to spend even more money to make the next real improvement to the value stream.
A value stream is not a department. A value stream is not a function. Very few places have only one value stream. More and more companies are creating what they think are value stream maps (VSM) to identify improvement opportunities and to visualize their process flow, but sometimes these VSM are merely process flow diagrams that use VSM icons. You cannot improve a value stream if you do not recognize the distinct differences between value streams and measure them separately. Optimizing a point in the system does not necessarily result in optimizing the system. Point optimization often creates an imbalance that degrades system performance.
All these problems had common root causes. They mistook a department as a value stream or failed to recognize the various value streams that went through common process steps. The conveyor in the packing department saved transportation, but the real opportunity was to eliminate packing as a department by combining picking and packing as one process step. The mezzanine improved flow and speed through the picking of small parts but required order consolidation, which created errors and delays that outweighed the savings generated by optimizing that one part of the value stream. The mezzanine also created a bottleneck that increased the total lead time and increased the costs for material handling and overall labor. Automating the weighing, wrapping, labelling, and movement of pallets of material from packing to the dock area was a big cost save, but not as big as breaking these functions into their individual value streams and putting them in the correct location in the facility to minimize or eliminate the need for transportation, handling and sorting.
Do not improve something that should be eliminated.
Cost reductions and efforts to improve cycle time can often cost justify themselves and show a return on investment (ROI), but if it does not improve the flow through a value stream and improve the system, then you have hardened an obstacle to real improvement, created a monument, and possibly added lead time and total costs. Do not automate waste and call it an improvement.
Robert H. Simonis is the senior operations consultant at KCE Consulting LLC. Robert has over 20 years of leadership experience, including 10 years in manufacturing management and 10 years of global responsibilities in automotive, electronics, machining, logistics and complex assembly operations and is recognized as a lean enterprise expert. Email [email protected] or visit www.kceconsulting.com.