Testing continues…

Posted by Leonard...

Written by The Lean Guy at Genoa...

When we set up this blog we had a couple of purposes in mind; document our journey and share our successes, failures, and lessons learned with others. As I mentioned previously we’ve collected reams of data from past projects and loaded all of it to an excel file knowing that somewhere in there was the path to establishing takt time in our design world.

Well, I’ve just completed a high level analysis of the data and have worked out some basic production averages including parts per hour and parts per volume. This is broken out by ship part regardless of type of ship and all project hours are used. There were some obvious outliers in the data so one standard deviation was used to establish upper and lower control limits to arrive at a more “reliable” average.

The data looked great in scatter diagrams with some obvious patterns emerging.

Remember, one of our goals is to use historical data to calculate project hours based on volume or weight or some other parametric. That said, we took the rates and tested them against a project that had already been completed, calculating parts based on the known volume of each unit, and calculating hours based on the calculated parts.

The results were a little disappointing with calculated parts off by as high as 85% but on average 20%. Calculated hours were off on average 33% but as high as 70% for one unit. I thought they would be better but have to keep reminding myself that this is data from all types of ships applied against one particular type.

A positive result of our analysis, which has shown to be consistent across all ship types and unit types, is the percentage of total project hours that can be applied against modeling, assembly, nesting etc.