Line Balance Rate

The operator balance chart, also known as a percent load chart, operator loading diagram, cycle time/takt time bar chart, or line balance analysis graph, provides the lean practitioner with insight into how equalized operation time is among the workers within a given process, line or cell. The line balance rate (LBR), and the related line balance loss rate (which is simply 100% minus the LBR), quantifies how well or poorly the line is balanced.

A lack of line balance routinely causes the waste of waiting and/or overproduction. It can also prompt over-processing during which operators, rather than engage in the blatant waste of waiting, conduct “apparent work.” Line imbalance is an enemy of continuous flow.

Some may ask, “What the heck do I do with this?” While there is not necessarily a magical LBR “bogey,” it’s definitely useful when developing standard work and comparing different balance scenarios.

Consider LBR a simple analytical tool. Use it when it makes sense.

line balance rate

There are 8 Comments

Dave's picture

Mark, in your opinion how would you use that 84.2% figure? If running multiple lines, could you compare the LBR% as a gauge? For example, is it fair to say something like "Line 1 is balanced at 84.2% and our target is 95% balance"?

MarkRHamel's picture

Dave,

Thanks for the comment/question!

Certainly, one may have a target condition that is say 95%. 84.2% definitely indicates that there may be some room for improvement.

In the realm of balancing operator load, the closer LBR is to 100%, the better, obviously. However, the ultimate metric is really productivity in terms of the number of units/person/hour.

Usually, the higher the LBR, the higher the productivity, given the same "playbook." Some playbooks, for example a 4 versus 5 person line, may be inherently more productive, because it lends itself more easily to balancing work content (until new improvements are made, anyway). I see LBR as a tool to help: 1) determine what playbook or staffing models are the least waste way, and 2) to (help spur improvement activities to first remove work content altogether and then) improve the balance through combining, rearranging, simplifying, and splitting.

I hope that's helpful.

Best regards,
Mark

Tim Sachtschale's picture

Thanks for confirming my initial approach with the shop floor - i.e. a "managed" buffer after the offending machine that is sized to the variation so as not to starve downstream operations.

Cheers,
Tim

Tim Sachtschale's picture

Mark - do you have any insights as to how the bar chart is best sloped from first operator to last? Are there situations where an upward slope is better than a downward slope, especially when buffering against machine variation that has no line of sight to improvement? Alternatively, how is the slope adjusted in conjunction with SWIP to maximize line capacity (flow) within inherent machine variation?

MarkRHamel's picture

Hi Tim,

Thanks for the question. Sorry for the tardy reply.

The conventional wisdom on the operator balance chart (OBC) is to load work content to the left, thereby loading the workers to takt time, actually, more appropriately, planned cycle time. The ultimate aim, after eliminating as much waste as possible, is to be able to redeploy the rightmost worker(s) who no longer have any work content. That of course is not always so cut and dry.

Usually the bar at the end of the OBC has some work content. This is where we typically apply some rules of thumb around optimal staffing such as: 1) if last worker is loaded less than 30% - do not include (add) an extra operator. Instead, reduce waste and incidental work. 2) if last worker is loaded 30% to 50% - do not immediately add extra operator. Run cell, line or work area for two weeks. Re-evaluate then if enough waste and incidental work can be removed to avoid adding an operator, and 3) if greater than 50% - add an extra operator if required. Keep reducing waste and incidental work to eventually eliminate the need for the operator within the cell, line or work area.

So, that's a long-winded answer that the slope is about equal to 1...until the last operator.

If there is significant machine variation that then drives cycle time variation (and no immediate opportunity for improvement), then your planned cycle time may (depending upon machine cycle time) end up being well below takt and thus you'll end up spreading the work over more operators. Or, depending where the poor machine performance is within the line, consider breaking the line's continuous flow (which is near impossible if you had large machine variation) and put some sort of buffer (FIFO lane or supermarket, perhaps?) in place to dampen the effect. SWIP, in the case of a FIFO lane would include the buffer inventory. I would not try to correlate slope with SWIP.

I hope this is helpful.

Best regards,
Mark

Deepti's picture

Hello Mark,

Could you please help me with Takt time and Line balancing problem, I am working in service industry. I have target of 208 reports per month, each report takes average 11 hrs to complete , assuming 7 working hours per day. How do I estimate the number of people required to complete the target.
Please let me know if you need more data.
Your help is very much appreciated.
Regards
Deepti

lloucka's picture

Deepti, how many days in a month? Lets assume 20 working days in a month.
Takt Time = Available Time / Demand = (7 hrs per day * 20 days per month)/208 reports per month = 0.67 hrs
# Report Writers = Takt Time / report writing Cycle Time = 0.067/11 hrs = 16.3 report writers.

Let's double check this ...
Demand per month = 208 reports * 11 hrs per report = 2288 hrs per month
Available Capacity = 16.3 report Writers * 7 hours per day * 20 days per month = 2288 hrs.

But ... the demand probably varies and isn't always 208 every month, and some reports are easy and some are harder, and some writers are faster or slower, so 11 hours per report might just be an average. So you will need more that 16 writers. How many more will depend on the variation in report demand, and writing.

Deepti Varshney's picture

Hello,

Thanks a lot for answering my question, you have really helped me a lot. Yes you are right there are some experienced analysts some are very new in the work. But the report demand is pretty stable. Also 11 hrs is avg time (process lead time). I am trying to reduce the process lead time to avg 9hrs.

Thanks once again for quick response and help.

Regards
Deepti