[trisampson]

Originally Posted by Marco

Hi,

has any-one ever used "Linebalancing" ? PLease could you give some insights how you performed and what the results were. Was it easy to implement? Especially when you have to re-allocate people du to the fact the line is more balanced and people aren't "needed" anymore for a "specific" task?

I know this post is long-overdue, but I thought it's still a worthy discussion and I didn't see much discussed about it...

Let's all make sure we understand that "line-balancing" is not the same thing as "load-leveling". Load leveling intends to level the aggregate production volume over an extended period of time so as to even the load on production resources.

Line-balancing, on the other hand, is balancing the standard work (in terms of operation time) within a mfg cell. This is a royal pain in the you-know-what if your demand changes frequently. If the demand changes, you may opt to add or remove operators from the line, which would require you to re-balance the line - and of course establishing new standard work for the number of operators you have on the line.

There are two ways to get around true line-balancing. The first is what I do in our facility. I "roughly" balance the line (I don't get scientific, I just try to get each operation close). Then establish a policy (using kanbans) that operators will stop their work if another operator gets behind. The operator who is behind receives help from the others and work continues. The operators can use their judgement on when to stop and help others.

The other, and better solution, is to use self-balancing lines. This practically requires your production operators to be standing (and therefore mobile). This method also uses TRUE single-piece flow. Speaking briefly, each operator continues working on the piece until the piece is pulled from the operator downstream. That is, say the last operator in the line finishes a part. That operator walks upstream and stops the previous operator at that exact time and finishes their work. The second to last operator then walks upstream to the next operator and initiates a pull until finally the first operator walks back to begin a new part.

Hope it helps...

[dchallender]

Hi

From a personal perspective I've carried out line balancing many times over the years and for it to work properly the folks on the shop floor have to do it rather than me as an Industrial Engineer for instance. In doing it this way you are empowering them to make the improvements themselves, I think the key to doing this as well is letting them make mistakes and learn from it.

Also line balancing is just one element of standard work, for example to balance a line you need to have a number of things in place to do it, such as

• Standard operations (Std Op) - understand the breakdown of the work content.
• Time and Motion data - detailed increments of time to carry out the Std OP - this translates ino takt time data and plannned cycle time, etc.
• Visual management & Control - ability to see abnormal conditions and control them
• Issue resolution and escalation chain - to problem solve and resolve the abnormality
• Flow management - kanbans, supermarkets, etc.

The process of balancing is relatively straightforward and when I have enough posts I will upload some info on it, however you basically use load balancing boards or to use the japanese Yamazumi boards. This shows the work content at each stage with a graphical representation of time using post its or magnetic strips. The strips are different colours for VA and NVA, showing where you can eliminate waste and where you can move work around. Again to stress this - the guys on the shop floor do all of this themselves, that way they will own it and implement it, once they have agreed on the work content at each stage - that thens becomes the new standard work.

I used this on a production line when I was in aerospace and asked the guys to estimate the times for each element of work as the sight of stop watches used to cause many issues. When they had finished estimating the times and the total work conent was added up they were 2 mins short of the engineering standard of 334hrs. This to me was a pivotal moment as it made me realise that given the opportunity to improve for themselves the guys on the shop floor will do their best' also from that day forward they used the balancing boards for themselves.

Finally on a question made earlier - what do you do if the rate changes?

Again if the rate change you should have already precalculated the work content and balancing for the various rates. For instance again in aerospace we knew the line configuration options we had to do for 3, 4, 5 or 6 units per week ( I know it all seems like low numbers - believe me in aerospace it's huge !!). To physically do this we moved the guys around the jigs on the line and they carried out different 'work packages' (Std Ops) for the various different volumes - however we'd preplanned all of this before hand as it was probable based on the demand analysis we had carried out. This is exactly the same prinicple they use in an automotive plant, the same production line but now we are doing different jobs, thats why in automotive there is such an emphasis on cross training.

I've done exactly the same in service industries however the variation in work content is a killer and so I have found that true line balancing only works with repetitive work, e.g. processing claims, etc. If there is too much variation then is is more akin to load balancing.

hope that helps and wasn't to much of a sermon

rgds

Darren

[Lakota]

darren,
Outstanding post. We look forward to your being able to upload the documents mentioned.

Looking forward to your many future posts. Your posts have been clear, direct, and provide value add and meaning yo fellow TreQnaites.

Thank you for sharing with us.

Blessings,
Lakota