By Jon Miller | Post Date: August 10, 2008 9:42 PM | Comments: 1
On a few occasions lately I've had to make the hard sell for cells. Typically when teaching people how to implement TPS we show more than tell. The learning by doing helps people take ownership of the decision to make a significant change. But more so than changing methods, making major physical changes to the layout of the process and changing the organization chart to match the new work flow takes some persuasion by telling and selling, before people are willing to commit and give it a try.
At the most basic level of process flow we have what are called "processes villages" where similar equipment is grouped together, managed and the productivity of these processes are optimized. The benefit is that the process villages themselves are very easy to manage, while the whole system becomes fairly complex. Typically this is the result of a small business growing organically by adding more machines or skills to the same area so that these resources can be shared, allowing any process to serve any customer. This apparent flexibility in fact creates complexity when seen from end-to-end rather than process by process.
Cells and flow lines are superior to disconnected processes, even with these disconnected processes are individually technically superior. The reason is that the connection of processes removes work in process inventory and other delays, resulting in shorter lead-times and improved on-time delivery. This allows the producer to synchronize output closer to actual customer demand in what is known as just in time production.
What most do not realize is that there is a diminishing return to productivity by scaling resources up or down on a conveyor-type line. At some point going 10% faster does not yield 10% more good product. When this happens the solution is to work overtime or to duplicate the conveyor line, both expensive alternatives.
In addition, conveyor lines with fixed positions of work typically do not allow for so much cross training and multi-process handling, due to how linear flow processes are set up with tooling, material presentation and work positions. Even when multi-process handling is promoted, there is a significant loss due to wasted movement of people back and forth along the line to operate multiple processes.
Converting large lines or process villages to a set of cells with the same total capacity can overcome these disadvantages. By having multiple cells (when appropriate) running at lower speeds, the losses to quality and safety can be minimized due to ergonomic aspects of slower paced work. While long conveyors lines may have long changeover times, stopping the whole line to change material or tools, multiple calls can reduce or eliminate changeovers, if the creation of cells along process families allows it. Both the initial investment and running cost of cells tends to be lower as equipment is designed to suit the smaller, slower process and "right sized" through production preparation try-storming. The running cost is lower due to reduced energy cost and better maintainability.
The major arguments against cells tends to be that more training will be needed for the workers to function in a cell, or that closer management attention and quicker problem response will be needed at the cell or team level as compared to the line or village level. To be blunt, if you are not willing to do the two things above in principle, you should stop pretending that you are implementing a lean system.
The other major and valid objection is that existing monuments (immovable equipment) or expensive shared equipment cannot be placed into the cellular flow. When this is true, the best approach is to first weight the safety, quality, delivery, cost, space and morale benefits against the cost of adding a piece or two of equipment strategically. When this does not pan out simply build the cell without that piece of equipment in the short term and establish what is called a tsurube system (empty pail goes down the well, full pail comes up) by setting up standard work in process (SWIP) within the cell and at that outside process, with firm lead-time turnaround agreements.
But I realized that credible arguments in terms of improved performance are simply not enough when making the hard sell for cells. At the end of the day, people can argue whatever they want based on their interpretation or understanding of numbers, claim "the risk is too great" to pick up and move a piece of machinery, or simply say "it's never been done in our industry." We need to appeal to a higher authority: humanity.
There is no doubt that stabilizing, balancing and connecting processes in order to make them flow is the right way to do kaizen. However the more important thing is that it is also the right thing to do in terms of respect for people. The cell brings processes together, thereby bringing people together. The basic unit for natural work teams to form work zones that connected physically and in terms of work flow. When the person working to either side of you directly depends on the work you do to be successful, team spirit can take root. Cells bring people together physically as well as in purpose. Cells are not the only way to build a team but they are a good way.
Speaking of process villages, I have never visited a village characterized by having all of the same occupation or trade type in a single place. Typically villages have a little bit of everything, or many small trades and shops so that the village can function and the villagers' needs can be met. A village full of just bakeries or blacksmiths? Perhaps in a section or neighborhood of a larger city but not in a village. A traditional village functions more like a cell when you think about it.Comments are moderated to filter spam and inappropriate content. There may be a delay before your comment is published.