Page 103 - Design for Six Sigma a Roadmap for Product Development

78 Chapter Three

4. Uneven job size versus similar job size. Again, we are comparing two
queue systems. In both queues, the arriving time patterns are the
same. The first queue system is such that every incoming job takes
about the same amount of time to finish by the server, for example,
10 minutes. The second queue system is such that every incoming
job has a different size; some jobs are big and take longer to finish,
while other jobs are small and take a shorter time to finish, even if
the average job processing time is the same as that of the first
queue system (for example, also 10 minutes). Then the waiting time
and queue length of the first queue system will be shorter than
those of the second one. The implication for the product develop-
ment is that loading engineers with similar job sizes for each task
will increase the job throughput.
The Toyota product development system (Morgan and Liker 2006,
Kennedy 2003) has developed some well established practices to
ensure smooth product development job flows. In Toyota’s mind, there
are three M’s that are very harmful in the product development
process:
1. Muda: nonvalue-added
2. Muri: overburden
3. Mura: unevenness

Toyota believes that in the product development process, muda is
the obvious waste to be removed. But muri and mura are another two
hidden wastes that should be identified and removed as well. Muri
means the overburden of design engineers; when design engineers are
under excessively high time pressure to rush through the jobs, they
tend to produce low-quality work with flaws. Mura means unevenness
in workload. For example, in many companies, on some days, engineers
work a 16-hour day, while on other days, they don’t have much work to
do.
Specifically, one of the key principles in the Toyota product develop-
ment system is to create leveled product development process flow. This
principle calls for synchronizing activities across different functional
departments in a product development organization. It also calls for
evenly distributed workload to various departments and engineers;
you will not want to see some extremely busy days mixed with
extremely idle days, and you will want steady loads for all the people.
This approach creates steady workload and job flow so the tasks will
flow through the organization smoothly and a waiting line will be
unlikely to occur. Even if you get a huge data set to be distributed to
other departments, you will cut the big data set into chunks and give
the people one chunk at a time.