I say first but it’s not really our first. Just the first that I’ve been involved in and the first since moving to our new plant.
Our factory, due to the diverse nature of the products we undertake is divided (not in the literal sense) into cells that work on particular contracts. This allows us to he the right tools in the right place for the right people to do the right job. For our first SCE, we decided to attack the cell that manufactures cables for one of our largest clients. After the build process, the cables are then delivered to the customer’s site where a team of our guys install the cables in flight simulators. It wasn’t done because the cell was doing a bad job or failing to meet targets. To the contrary, in fact, the team were already doing a great job producing the goods on time with very little re-work. Some of you that are reading this may be thinking along the lines of “if it works already, why change it?” or “if it ain’t broke, don’t fix it”. Quite simply put: Anything can be improved. That is what Lean is all about. Continuous improvement!
What were the goals?
To eliminate costly waste in the process.
To reduce lead time.
To work towards a Pull system.
To lower the amount of WIP.
So what did we do?
First of all, we got a team together, consisting of people both inside and outside of the process to brainstorm out every step of the process and gather a few ideas about what we can do to make it better. We wrote each operation within the process on a post-it note, put them on the wall in the order that the current process works. This was a nice visual way of seeing the whole process, end to end. Once we had done this, we identified which operations added value and which were waste (everything that does not add value is essentially waste). We used a different colour post-it for the value adding operations. This was a real “Eureka” moment for the team. They actually started to ask questions at this point. The team’s points and questions were also put on post-it notes for use later.
Using the data collected from the team meeting, we walked through the process, creating a string diagram showing where and how far the people and work move through the process, also noting times taken for each step and/or operation. We worked out that in the process of making the cable assemblies for one simulator (approx 750), our production staff walked around 150 miles between them! Another thing that was evident was that at various stages of the process, the assemblies moved around in batches. The entire system was a traditional push system. The overall lead time for one simulator was roughly 7.5 weeks. As I’m sure you can imagine, the amount of WIP kicking around due to shortages and Engineering Change was staggering.
What did we do to improve it?
First of all, we established that there was a lot of checking going on so we removed a lot of the checks done throughout the process and put in place visual measures and tools to help the team get things right-first-time. We also re-arranged the shop floor so that the flow of the materials through the process was linear. We moved common tools nearer to where they were being used and made shadow boards with photos of the tools on them so that everyone could see what tools went where. We changed the old batch push system for a single piece flow, Kan Ban pull system. In the old system, the jobs travelled through the process in zip-lock bags so there was a lot of bagging and unbagging throughout the process. The bags went out the window and were replaced by boxes. We moved inspection into the process. Previously they were in another area on the shop floor. We also applied some 5S throughout the system (this included things like painting bench tops, using marker tape to show where things should be etc). Our cable cut area could be used as a best practice example!
What did we gain from it?
The lead time for a simulator is now approximately 2.5 weeks (that’s 5 weeks removed!)
The 150 Miles figure is down to a few hundred yards.
An output rate of approximately one cable per 7 Minutes (because of the batching it was the same as the lead time for a whole simulator).
Much lower WIP.
What is left to do?
In short, a lot! Remember what I said about continuous improvement? Perfection is something you will never get but all the while you are moving towards it, you are doing the right thing. We still need to look at standardising the workstations and removing a lot of personal clutter. There are still some issues with the amount of checking and data collection that need to be addressed. We need to address level loading of work through the process. The cable assemblies are wildly different in some cases so that can and does occasionally cause wait time in the process.
As you can see there are still plenty of opportunities there to drive improvement and no matter how good we get, there always will be. Once we’ve crossed the t’s and dotted the i’s on that cell we will leave it alone for a bit as far as change goes but re-visit from time to time to collect data and improve further.