21 Questions to Ask When Walking the Model Line, Part 1

The model line is a value stream or a section of a value stream used as a pilot to demonstrate an organization’s capability to deploy lean systems and behaviors. The model line approach is used early in an organizations lean journey as a way to achieve rapid results in a focused area, learn about both lean and what it takes to implement, and to prepare for wide-scale lean implementation. A good test of a model line’s maturity is to ask one or more people with deep expertise in lean to conduct an audit. Two full days are recommended even for model line of limited scope, to allow for enough time for several gemba walks with management, informal interviews of support staff and people on the line, and the presentation of a summary of findings. If pressed for time, and wanting a quick guide to gauge the maturity of your model line, here are 21 questions you can start with:

1. For how many consecutive days was the required demand met within regular hours?

The answer may seem to be “the longer the better” but in fact a 100% rate of operation suggests plenty of slack or hidden problems. The target should be around 95%, or perhaps that one day in twenty required overtime. Taiichi Ohno taught us that the line that never stops is either really great or really bad. Whatever the answer to this question, it should yield hints of the areas of losses on the line and point to other areas of questioning.

2. What is the status of the production for the hour against plan?
One of the first things you see on the model line should be an “hour by hour” chart of hourly production control board. A digital version simply will not do. At an advanced stage it may be allowable, but when first implementing lean the manually updated white board or flip chart production control board insures local ownership and engagement. Be sure to check on the freshness of the information and that where problems have been identified, the team has thought through to the root causes.

3. How do teams on the model line begin their shift?
Whenever possible, arrive early enough to observe the shift start at the model line. This should not be an uncoordinated and unsynchronized startup but rather a sharp “ready-set-go” process. At a minimum a world class model line would have a team meeting at the line or the visual boards near it, some brief checking and tidying of the workstation, and the simultaneous start with pre-set in-process goods on the line.

4. Where is the pacemaker process?
The pacemaker is more than a mechanical point for setting the speed or “pacing” the line, it is the single location where scheduling takes place. The remainder of the line should be triggered by the pull signal from the pacemaker. Too many model lines operate on quasi-pull due to lack of a clear pacemaker.

5. How are “islands” or processes set apart or isolated from the upstream and downstream processes connected?
A model line need not necessarily a series of equipment in a contained area or even a physical production line. The value stream may span several departments and be physically disconnected. While this is not ideal, rearrangement of equipment is not always a practical first step as part of a model line project. In fact the moving of equipment can become the end in itself, even taking resources away from training or improvement to the equipment in situ. It is better to ask how standard WIP, the pulley system (tsurube) and other methods are used to pull between islands.

6. Where are the standard work documents?
The standard work chart(s) should be visible. Hunting and retrieving standard work documents from their storage place is far from world class.

7. How well is standard work being followed?
This is a question rich in possibilities. Is the standard work easy to follow? Does it appear easy? Is it in fact easy? Why or why not? The better and more detailed the standard work, the more ways this question can enlighten us as to the maturity of the model line.

8. Does standard work exist for occasional work such as tool changes, material changes and the replacement of empty or full containers?
Many times problems are hidden in the occasional or filler work. When this work is timed and standardized it becomes possible to improve it. Another way to ask this question is, “What has not been standardized and why?” This one is a lot of fun, but can put the questioned in an awkward spot. It should be used only when there is enough trust and rapport with those people.

9. When was standard work last updated (improved)?
The smaller this number the better, as long as the unit is never more than days. Twenty one is a good initial target for a mature, stable model line.

10. What percentage of balance to takt time has been achieved for manual cycle times for individual stations?
The model line should display a yamazumi chart that shows stacks, like wooden blocks, the individual work elements at each station. Against a horizontal red line of takt time, some stacks will be naturally taller than others unless perfect balance has been achieved. An overall balance to takt time of 95% is ideal but for mixed model production 10%-15% is common. Also look for kaizen activities generated from imbalances on the yamazumi.
Jump to Part 2

5 Comments

  1. Joseph

    June 9, 2010 - 4:26 pm

    Jon.
    I can not wait to see 11 – 21. Now we are talking proper LEAN. I would not agree with a 95% loading that is stated in step 10. It is too slack. Workers can raise their effort by 10 – 15% quite easily for quite long periods. The peaks will balance with the troughs. With model mix / variant work weighted average balancing should be done to balance the work load at 99.99% on all stations. I would have the Team set the longest cycle to be no more than 10% above Takt. All critical operations must be completed within the first 70% of the station for QC / Safety reasons.
    With Lean I would expect the Group Leader to over see and assist any away from standard situations. Toyota make a case for the last operator to have any slack. This is so you can work towards removing that operation. I would expect 2 sections to have 1 station of this nature.
    I totally agree with your statement in step 8. All standard, cyclic & non-cyclic operations must have times & WES sheet. You must manage so the line runs under all conditions. Most none cyclic work can be done in the troughs of the line balancing. This should all be shown on the Yamazumi board. Use Black for these away from standard operations.
    Being Lean lowers the water level but doesn’t it make life more interesting. I would never, never let an operator have an allocation that he could not do at a correct work rate. It would make no sense.
    If any allocations do not work then the TEAM would resolve them. All allocations should be done by the TEAM. They do not have the keys to the family jewels though.
    In steps 1 – 2 if the line runs consistently the biggest challange is to keep the operators interested. It becomes like “Ground Hog Day”. The team should make this their priority.

  2. Jon Miller

    June 9, 2010 - 5:22 pm

    Hi Joseph
    Thanks for your in-depth comments around yamazumi, standardized work and kaizen by the team.
    Balance to takt time of 99.9% is pretty hard core! Keep in mind the model line in this context is used by companies that are approaching lean but not yet running for years at Toyota’s level.

  3. ericmo

    June 9, 2010 - 8:02 pm

    Joseph,
    At the end of any Kaizen event for this case the model shop, a 95% balance to takt percentage is the appropriate measure. It is the setting of metrics value that I think is being pointed out and not the perception whether operator is efficient or not against the measure. However, in kaizen perspective, it is always expected that productivity improves after some time of continued practice and so I agree with expecting a 10% improvement in productivity as the target for a follow up kaizen.
    Setting CT at maximum (99.99%) against takt time will not give you a buffer for occasional problems that’s likely to occur e.g. sudden part supply delay, broken tools, etc.
    Critical operations may also depend on the assembly process sequence which is impossible to change or relocate due to physical or functional constraint in the assembly specification.

  4. Paul

    June 13, 2010 - 1:52 am

    Agree with 95% balance on the Model Line. Toyota will load up to the 100% for most type 1 processes. They are mature enough to do this as they have already eliminated/reduced the 5% fluctuation within the process, i.e. replenishing parts – this will be given to a type 3 person who has standard work to “fill” all processes up. This reduces fluctuation but also reduces variation in the type 1 work – leading to better quality as the standard work is not broken. No abnormality % is factored in either, this will make problems visible if they should occur.
    As for the slack in the last processes (flexible pitches) this was used in the main for rebalancing of work due to takt time changes. Toyota have now moved away from this as it was found numerous defects were being created on this process, i.e. hoses missing (relatively few interlocks on the pitch).
    As most processes within Toyota will be type 2 (derivative fluctuation) you won’t find many that are not averaged out to the TT line.

  5. Joseph

    June 21, 2010 - 12:23 pm

    Paul.
    Many thanks for your info. on the Toyota allocation system. Having never worked for them I only have old second hand information.
    I have watched the video on “Toyota Trim Lines” on the other blog. I could see that the lady working in the Engine Compartment did not have any time to re-stock her Yasaki Trolley. It goes against my back ground to have operators just stocking up with the associated waste that their job would have. It must work on “Total Cost” or Toyota would not be doing it.
    As Toyota have a Group Leader ratio of 1:5-6 operators I would investigated having the GL stocking up his team or some part of his team.