Finish Systems Engineering before starting Production

Having worked in Systems Engineering for many years, I have seen the mistakes made by Management for many times. An order would arrive at the business, and production would start to build it. Systems engineering would not have carried out an analysis to find out what was wanted.

A week later, systems engineering would probably get sight of the contact and start to carry the analysis of what was required. The detailed analysis would continue for about six months, and systems engineering would publish the design report. Gradually the system engineers would speak to production to find out what they were building. A month later, change would start to the design; this would continue until the delivery went out of the door.

Change is cheap when a project is in its infancy, but as the project matures, the cost of change grows exponentially. At the end of each project, there would be an inquest of what went wrong. Management would make recommendations to wait to start the production until systems engineering has completed their analysis. The next project would arrive, and production would start it. For a while, this went on until systems engineering got upset and demanded on the next project that production should wait until system engineering is finished their analysis.

The next project arrived, and this time production did nothing, and systems engineering started the analysis. Production phoned almost every day to enquire if systems engineering was finished. After two months, out came the analysis and the design. Production built it from the design, not from what they thought it should be.

The amount of change on the project was next to nothing. Not only that, the unit was delivered three weeks early. The profits from the project were huge. A company's dream to have units going out early next to no change, and the customer delighted.

Doing systems engineering first is not rocket science, but it is the most logical thing to do. Change is not good with any project, but if you can make the changes first early before any metal has been cut or code written, it is ridiculously cheap to make changes to the project. You delete a line on a drawing and add a new line. Make the same change just before the unit goes out of the door; the change must be made. The unit retested, documentation updated, the change will have to be managed; this becomes expensive and takes time and effort. The customer does not like to see change late in their program.

Carrying out the systems engineering analysis before starting to build has far more advantages than building and carrying out the analysis simultaneously.