Predictive Problem Solving

I answered a question on a forum earlier this week, and it has had me thinking. It asked how we thought efforts to improve quality would be different in the future.

My answer was that I saw a more predictive approach to quality. That just means that more companies will make the effort to identify the drivers of poor quality, and track those, rather than put all their resources into identifying faults. For example, perhaps a high percentage of inexperienced workers is a leading indicator of increasing defects. Or poor quality could be related to weather, or to passing a capacity threshold. In short, I see people getting smarter with the way they manage quality.

Predict Problems Before They Happen

The same predictive problem solving approach applies to any aspect of Lean. Quite a while ago, I did a study trying to find a link between absences and quality. My premise was that when many people were out sick, quality would drop. Instead, it got better. Why? Because there was more line stop due to the inexperience, which gave the majority of the team some extra time to check their work.

That sort of insight opens up a whole range of questions. Is the Standard Work a problem if the normal pace generates poor quality? Are there flaws in the cross-training plan if the backups are ineffective? Is there a pattern to absences that could be addressed?

Predictive problem solving is the act of looking for the indicators, which are closely tied to the root cause of a problem, and stopping problems before they happen. Of course, that takes good data collection plans, daily management, and a curious mind to ask a lot of questions.

I am sure there are some of you out there who are already tracking these sorts of leading indicators. I am equally sure my readers would appreciate your insight into what you are watching to nip problems in the bud. Care to share?