Working with multiple methods

- One of my clients, who works in the automotive industry, had a recurring problems. A robotic welder kept crashing against the piece it was supposed to be welding. The impact would bend the welding tip, causing the machine to make welds in the wrong places. To find the root cause, the team first went out to the production floor to learn the complete process, from setting up the automation, until the finished part was delivered. They walked each step to observe, map the process, and ask the operators questions. Then they drew a fish bone diagram that showed several potential causes of the malfunction. They noted factors such as inadequate operator training, incorrect programming of the machines, the impact of the extreme accumulation of heat on the welding tip, and the incorrect assembly of the tip into the automation. Since the company was not tracking the data of rejects, there was no Pareto analysis, and the team used a set of questions based on is, is not, to narrow down the most likely causes of the problem. What is happening? The welding tip is crashing against the piece it's supposed to weld. What is not happening? The automation is running correctly. So the automation is not the problem. When does it happen? On second shift. When is it not happening? On first shift. Where does it happen? When the programming steps are not followed correctly and at the assembly of the welding tip. When not? When the programming is done correctly and the tip assembly instructions are followed. By asking these questions, the team narrowed down the root causes to a set of instructions not being followed by the second shift team, and they proceeded to use the five why to determine the deeper root causes. When using the five why's, it's best practice to start with the original problem. The part is defective. Why? Because the welding tip is crashing into the part it's welding. Why is the machine doing that? Because the operator did not assemble the tip correctly. Why not? Because they did not follow the latest operation instructions. Why? Because the second shift didn't receive the latest training. Based on this answer, the team agreed they could correct the specific problem by making sure every shift is trained on the latest operating instruction, which, by the way, corrected the problem. However, the team continued asking more why's. Why were the second shift operators not trained? Because the company relied on one person to do the training, and that person only worked on first shift. Why rely on one person? Because the company had no formal procedure to ensure training on every shift. By implementing a formal procedure, the company took care of other similar problems. As you can see, while no magic formula exists about when to use one tool or another, often by combining several tools, it's more likely that your root cause analysis will be successful. Using many tools can help ensure that the cause and effect chain makes logical sense, so that you're uncovering the real root causes and you're countermeasures can effectively prevent your problems from recurring.