So during a recent conversation with Duane Siemen, a reliability engineering manager and classic "tweaker", we were discussing process improvement. Duane was sharing some of the work that he had done in the past to improve system reliability and throughput by focusing on the controls, communication, and tweaking how each machine in the system interacts with the system as a whole. I could not help but to compare the work that he did to the work of some of the great car tuners like
Hennessy,
Stillen,
Roush, and
Lingenfelter. When you think about cars or process equipment a few things are true for both. When you buy a stock car or a production line it is tuned to be conservative and somewhat mediocre. The car is not tweaked for the way you drive or where you drive from an environmental stand point and it is always "value engineered" to control production cost. It is set up to more or less work everywhere but excel nowhere. This means that without tuning and an understanding of operating context you leave a lot of performance on the table. Your production line is ordered with equipment from different manufactures that may or may not be set up to communicate with each other and may or may not like to run at the system rate. Many of the elements of the system may have been just dropped into the line without any concern for the up and down stream equipment in the system. Designers some time work under the philosophy of "will it work"not "will it work best." Did that designer know all the specifics of your widget packaging process or did they just create a packaging line?
So what if you wanted to tune your production line like Hennessy tunes a Viper.
The first step would be to understand operating context. A drag car is completely different from a track car.
- What are you using the line to do?
Second, learn all the equipment in the line. Without a full understanding of the dynamics and capability of the components then you can not begin to tweak. Roush would never take a car they had never studied and start changing ECU settings. They take the time to understand the parts before they tune the whole.
- What is the full capability of the equipment and parts in the production line?
Third, we focus on getting the system unified through communication. We want each piece of equipment in the line to have solid communication with the others so that each can tell the other parts what is needed to push performance. Some parts of the line may actually slow down to improve quality and this in turn will increase overall output but only if they can communicate their need to us or the system. In racing they say, "you have to slow down to go fast" that is what we want to be able to do with the equipment.
- Are the parts talking reliably with the whole? Are they using that information to improve system output?
Forth, we begin to problem solve and truly tune. We change a parameter and ensure we get the response we predicted. Does it effect the whole positively. We then use root cause to understand when it does not.
- Do the changes give the results we expected?
Lastly, we focus on sustaining the changes. As you tune and increase performance tolerances get tighter. Precision become more important. Operating context needs to be stable or at least understood. Because of these factors you may need more built in checks and possibly more preventive task to keep the equipment in spec and operating at peak performance. You may have to go so far as to build a daily management plan to maintain the levels you expect to produce the performance you want.
So in the end you need to understand your conditions, capability and components to get from minivan performance to race day ready production.