Rethinking Resource Sharing

Report
Rethinking Resource Sharing
Edward D Rothman
Professor of Statistics
University of Michigan
W Edward Deming
• 1900-1993
• American Statistician
• Contributions included a transformation of
the Japanese automobile industry from low
quality to high quality
• See the white paper “If Japan can why can’t
we….” 1981 on NBC
System Thinking
• A collection of components that come
together repeatedly for a purpose
• Our job is the achieve the purpose
• This requires an ability to plan to meet
customers needs and wants and to act
• The key is a useful theory and action based
on the theory.
Why do we need this?
• Many standard approaches to management
place emphasis on aspects of the system.
• For instance, measurement, MBO, and work
standards are designed to make us more
efficient.
• Principle: High efficiency may not be an
effective way to move us forward.
For Anne Gregory-W.B. Yeats
•
•
•
•
•
•
•
Interrogator:
“Never shall a young man,
Thrown into despair
By those great honey-coloured
Ramparts at your ear,
Love you for yourself alone
And not your yellow hair.”
And Anne Gregory replies,
•
•
•
•
•
•
“But I can get a hair-dye
And set such colour there,
Brown, or black, or carrot,
That young men in despair
May love me for myself alone
And not my yellow hair.”
Interrogator Responds:
•
•
•
•
•
•
“I heard an old religious man
But yesterday declare
That he had found a text to prove
That only God, my dear,
Could love you for yourself alone
And not your yellow hair.”
Example
500/hour
200/hour
Customer
Optimization Means Achieve the
Purpose
• In the last example, we want to have high
throughput yet little work in process (W.I.P.)
• Operate step 1 at 40% efficiency.
• Look at the rate limiting step---move this to a
higher level if the customer demands more.
• Recognize that those steps that do not limit
throughput, have excess capacity. What other
uses can we make of step 1?
Purpose
• The purpose of the system is an assertion that
represents a win for everyone.
• By everyone we mean suppliers, people in the
process and customers.
• We measure success by looking at how we
move towards achievement of the purpose.
• Principle: Optimization of a measurement can
be suboptimal for the system.
First Steps
• Identify the participants
• Learn the needs and wants of each party
• For each participant, ask why again and
again…
• Find the purpose
• Example: Your 14 year old daughter wants to
attend a party at a friends house. Seeks
permission to return at midnight. Dad wants
her home at 10 PM.
Mission=Purpose
• To (expand and promote---to provide
information---) information accessibility to the
…….community
The Learning Process
• The system design must recognize variation.
Seek robust methods—those that work well
under a wide variety of circumstances—but
also understand the need to adapt and
change.
• Deming Learning Cycle-Plan-Do-Study-Act
• Rothman-Rabkin: IPCAR—Imagine, Predict,
Choose, Act, Review
Voice of the Customer
• Many systems fail when we allow the least
informed to move us from step to step.
• The customer usually knows least about the
process.
• However, they can describe features of the
ideal process. (e.g. wait time for service)
• What features of the process can they help us
understand?
Principles
• In responding to the voice of the customer
make sure we understand the implications.
• Ex. A hospital clinic adds examining rooms to
reduce patients wait time in the clinic
entryway.
• Use a Pareto chart to identify the most
common issue or the most painful issue.
Work on these priorities first.
Voice of the Process
• You define the process and should highlight
features that are important to the customer
• Language is a key element—
– You need operational definitions of all
measurements, there is no true value
– The language you use internally may differ from
the customers language
• Measurements are used to focus attention on
aspects of the process
Some Principles
• Attempt to move upstream in the process
rather than focus on symptoms. (Ask why
again and again…)
• The solution is often at the interface of the
units, people, departments…. and not within
these areas.
• Technology is nice but often opportunities are
found that cost little and deliver much.
Measurement System
• Must be aligned with the process purpose
– Consider that a grade point average creates a
distinct purpose –
• The measurement is a model of the system—
just as a photo is a reflection of only aspects
of the subject.
– It is possible to improve the measurement at the
expense of the process.
– When an improvement in the measure is
observed ask “What was done?”
Measurements
• Outcomes such as time to deliver
• Predictors—Outcomes are too late and too
expensive to change—we need to move
upstream to make improvements
• Time –order-spatial data to learn about
stability
• Stratifying variables—related to outcomes and
to the groups compared
Time-Order
• Most systems are such that even when the
inputs are fixed, the outputs vary. For
instance, not every request will be met in
exactly the same number of hours.
• When the proportion of outcomes of a certain
type remain the same from day to day—the
process is said to be stable.
• The question you ask depends on whether a
system is stable!
Common Cause
• When a system is stable ask- What is the
common cause of the variation?
• When a system is unstable ask—What is the
special cause of the event?
• Example: A child spills milk…..
• The way we decide whether a system is stable
is by plotting the values against time or order.
Stratifying Variables
• Average days to completion of request for two
libraries:
• A: 3 days
• B: 4 days
• Is it possible that B always provides faster
service than A?
Strata
• Stratum I
• A: 2 days proportion 5 of 6
• B: 1 day proportion 1 of 4
• Stratum II
• A: 8 days proportion 1 of 6
• B: 5 days proportion 3 of 4
Learning Culture
• Will Rodgers once said; “even if you’re on the
right track, you’ll get run over if you just sit
there.”
• Beware of leaning disabilities –”I am my
position”, “the enemy is out there”, “fixation on
events”, “politics makes sameness an asset”
• Don’t engage in wack-a-mole
• Avoid benchmarking, numerical goals without a
plan, and competition between units (people)..
Thanks You
• Questions?

similar documents