Thursday, August 7, 2014

Quality and Acceptability of Solutions

In the past we have explored the difficulty of change.  One concept attributed to General Electric that we can wrap our heads around is Q x A = E.

Q = Quality Level (1-10) of the Solution to Solve the Problem
A = Acceptance Level (1-10) of the Solution in the Organization
E = Effectiveness (0-100) of Deployment of the Change

Let's look at some examples...

1) The solution is AWESOME!!!  It will solve the top two defects that make up 88% of the overall problems being experienced on the main product line.  On the Quality scale it is an 11.  Unfortunately it will require the replacement of three machines that were purchased four years ago and the capital costs will never reach even half of a return on investment.  A replacement material was identified with better properties, but the supply chain cannot support the required output.  No one is remotely interested in this solution and scores a 0.

In this case we multiply 11 x 0 and somehow we just intuitively know this idea is a big goose egg.  No way is anyone approving this solution.  In this case the Idea Fairy is not your friend.

2) The solution is just OK.  It will address most of the problems, assuming everyone follows the process without deviation.  On the Quality scale it is a 5.  The team has reached consensus about implementing this process update, even though the team agreed on moving work from over-loaded workcells to those with excess capacity and not everyone was super-excited about it (guess who...).  But they all decided it was the right thing to do and it scores a 6 on the Acceptance scale.

As before we multiply 5 x 6 and this scores a 30 out of 100.  It is certainly better than the solution from the first case study, but still fairly low for Effectiveness.

3)  The solution is good, not perfect, but good enough to address nearly all the problems with the top defect that cause the most rework and schedule disruption.  On the Quality scale it is an 8.  The team performed some radical brainstorming in small groups then brought the ideas together to merge the "best of the best" and then simulated the flows with balanced cells in the flow.  Everyone on the improvement and management teams feels absolutely confident that the solution will improve long-term quality and cost performance.  This idea scores a 9 on the Acceptance scale.

The product of 8 and 9 results in a 72 of 100 on the Effectiveness scale.  I think at this point there should be a pizza party planned and then starting the implementation of the idea.

This study addresses the importance of not just having one potential solution and the team should attempt to develop a suite of potential solutions.  Another good point is to not let Perfection get in the way of Good Enough.  Systemic problems respond differently compared to individual issues that can be categorized as Special Cause.  Socializing the solution through the value stream can help to increase Acceptance, but be ready to answer any question about the potential solution.

What kind of nightmare solutions have you seen rammed through implementation?

2 comments:

  1. We had an in house First Time Quality (FTQ) issue, directly attributable to a purchased part.but instead of making the supplier fix their issue (we're new, we took this job on as a favor to you) we had an in house process designed to do something entirely different, and we used that to mask the problem, and pass the internal FTQ test. Nice. 2-3 months out in the field, and the former FTQ problem now manifested itself as warranty. The final customer were ticked (surprise) and we caused a serious breach of trust with the intermediate customer. I wonder what would have happened if we had done a robust FMEA..........

    ReplyDelete
    Replies
    1. That certainly qualifies as a "nightmare solution". Looking at risks with the new process may have identified the resulting issues, and having an FMEA be robust would need the right mix of stuff-knowers during the team session. Having a history of past improvements that can be queried by part, process, or defect may be another way to increase the likelihood of a positive outcome.

      Delete