In my time spent onsite with the customer implementing PFEP (Plan-For-Every-Part) and advanced material flow techniques, I often was pulled into other projects. One of these projects was an effort to increase productivity (labor effectiveness, specifically) in a series of sub-assemblies. We were working on a single, mixed-model assembly line which was scheduled with a production (heijunka) wheel. Heijunka is defined in the Lean Lexicon, Third Edition as:
"Leveling the type and quantity of production over a fixed period of time. This enables production to efficiently meet customer demands while avoiding batching and results in minimum inventories, capital costs, manpower, and production lead time through the whole value stream."
As such, each day was planned to be very similar with regards to product mix coming down the assembly line. As in many real-world situations, reality didn’t always match the plan. The biggest deviation from the production wheel occurred with the lowest volume finished good. Another difficulty of this model was the work content. It contained nearly 50% more work-content than the model with the next highest work content. The remaining models on the assembly line were similar enough in work-content to balance effectively with scheduling alone. In order to balance the high-work content model, it was spaced in the production schedule as much as possible.
Let’s put some numbers to this scenario to better understand. For purposes of discussion, let’s say the assembly line produces 100 units per day and let’s call the low-volume, high-work content model the “Vortex." The Vortex was produced at an average of 10 units per day, but saw volumes as low as 3 per day and as high as 20 per day. To balance the work-content, four sub-assemblies produced four assemblies which were installed on the Vortex. Each of the sub-assemblies was staffed by one to two operators, depending on the volume of Vortex in the mix. As you may have already guessed, the problem with this situation was staffing. At low volumes such as three per day, four operators were working the sub-assemblies, but were waiting around quite a bit. At high volumes such as 20 per day, six or seven operators were scrambling to produce the assemblies. There were sub-problems also, such as the variable takt time of the sub-assemblies. With 20 Vortex models in a day, the takt time would be 24 minutes while at three per day it would be 160 minutes.
Our solution, which came to us after several planning sessions and some quick-win material flow improvements in the sub-assemblies, was the “rabbit chase.” For those of you unfamiliar with the rabbit chase concept, otherwise known as operators-in-motion or caravan, it is a work-cell design based upon the U-shaped work-cell. Instead of one operator in the work-cell, however, multiple operators “chase” each other through the loop, as depicted in the representation below (image credit to Dr. Ken Kirby of University of Tennessee). Through a series of planning sessions, we laid out the steps necessary to set up the rabbit chase layout:
- Collocate all four sub-assemblies.
- Break out and balance all work tasks.
- Layout inventory, tools, workbenches in such a way to complete the tasks sequentially and linearly.
- Define and set in place Standard WIP and Standard Work.
- Standardize spacing of Vortex models in the assembly line scheduling.
- Determine how many operators needed to satisfy spacing for different concentrations of Vortex models in the mix.
- Train all operators in Standard Work.
Above you can see a representation of our implementation. One challenge in collocating the sub-assemblies is that some workstations inevitably move further away from the install point of their output assembly on the line. We countered this challenge by using a layout that facilitated a means for the output assembly to “catch a ride” on the existing replenishment tugger route.
We achieved many benefits after this implementation. The biggest benefit was the ability to change the volume of sub-assembly output by simply adding or removing operators from the rabbit chase. This allowed us to match our labor requirement with the work required and remove vast amounts of operator waiting time. Another benefit is the cross-training the operators received. Each operator felt special to be a part of this new technique and was excited to learn each new task. The rabbit chase also instills a sense of competition to maintain pace in the loop. Since each operator is following the next one, you also prevent the operators from running into each other as they would in the U-shaped cells. Finally, another great benefit was the natural control of WIP. Since operators can’t pass other operators in the loop, they aren’t able to build up WIP.
While we saw many benefits to the rabbit chase, we were not without challenges. One downside to the implementation was the cross training I mentioned as a benefit. While the operator may have been excited to learn more tasks, we were challenged not to overburden them with too many tasks. Also, implementing the rabbit chase with the right balance of tasks can be difficult to do. Balancing is more effective if tasks can be broken down to true elementals. In reality, however, some tasks cannot be broken apart from other tasks. It is for this reason that timing within the loop is difficult. One imbalanced task too close to another imbalanced task can cause the spacing between operators to become more and more imbalanced until every operator is waiting on the first one. Slow operators can have the same effect. Finally, the collocation of sub-assemblies may have placed a workstation in a location that was not ideal.
In our case, the benefits of the rabbit chase implementation greatly outweighed the disadvantages. The challenging task was actually much more complicated than my simplified example and took about 6 months to complete. It was a great learning exercise and uncovered many process problems that we may have never seen otherwise.
Written by Kelcy Monday, Lean Deployment Specialist at LeanCor
Posted by LeanCor Supply Chain Group
LeanCor Supply Chain Group is a trusted supply chain partner that specializes in lean principles to deliver operational improvement. LeanCor’s three integrated divisions – LeanCor Training and Education, LeanCor Consulting, and LeanCor Logistics – help organizations eliminate waste, drive down costs, and build a culture of continuous improvement.Facebook LinkedIn Twitter Google+