Mean time to incident.
I’m working on a lean sigma project and I want to improve near miss and good catch reporting. My issue is using data to show it needs improved. How can I measure or show how many near misses I should target ? I was thinking about using the incident triangle but Heinrich’s data is known to be falsified. So how can I estimate or show how many I should target ?
Comments (7)
Yeah, disregard the Heinrich Theory. I would just take last year's totals and try to do X% reduction (10% less, 25% less, 50% less, whatever) - make it something that you have to work for, but is attainable/not out of reach. Otherwise, you're just making up a random or irrelevant number. Just keep in mind that near misses are lagging indicators (as it pertains to how most people track them), so you don't want to "target" too many.
I meant to answer this last week and never got around to it.
Lean and six sigma are methodologies for improvement. There are different improvement tools for each, with each having its pros and cons. Whether you're using A3 or DMAIC (or DMADV), the key to success is to follow the process. Both processes are best accomplished as a team rather than by one person.
For either method, the first step is to define the problem. In your case, you want to improve near miss reporting. In my organization, I might encourage the person looking at this to take a step back. Is improving near miss reporting for the sake of improving reporting the goal we really want to target? You say you struggle with using data to show it needs to be improved...if you don't have data, why do you think it needs to be improved? Start there.
It seems to me like you've already selected a solution (near miss reporting / hazard recognition) without defining the problem.