Page 1
- Welcome
Page 2
- Update
Page 3
- News
Page 4
- Brief lines
|
* Welcome *
Welcome to 2PLM, an e-zine distributed about every two weeks. This issue includes :
- 80-20 PLM
- The Evolution of PLM Theory
- Global PLM 2007 Survey
- PLM News
* 80-20 PLM *
By Menno Huijben
Mr. Pareto, known for his 80-20 rule, shows up in every PLM initiative. He appears in cases like these:
- "How much effort do we invest in this pre-study?
- Well, if in two weeks we get a good enough assessment (80 %), why spend 8 weeks more?"
- "80 % of what the PLM vendors provide is
the same. Let's spend 80 % of our evaluation time on the 20 % difference that matters"
- "We will be in pretty good shape if
we get 80 % of our business requirements covered with this small budget"
When PLM project deadlines draw near, and budgets run out, the original scope
of delivery is usually attacked with the 80-20 rule: better to deliver
80 % now than the full 100 % much later.
Reducing the delivery scope with the 80-20 rule works well when applied to a PDM system's features
and functionalities. Balancing manual process steps versus system transactions,
80-20 can be adopted as in this case:
- "In this release you still have to manually transfer
the CAD export file and check it in over there, but all
stakeholders will be automatically informed"
This example saves us the integration and testing work of the CAD interface (doing only
the bare minimum 20 %), while it gives us 80 % of the benefits by having
the CAD files vaulted and easily accessible for other designers.
The above examples are all very reasonable management decisions based on Pareto's
Principle. However, when dealing with (product) data quality, beware of the
80-20 approach. Consider the following:
- A report containing only 80 % of the data is incorrect
- An ECO (engineering change order) referring to 80 % of
the affected items can cause a lot of confusion
- Supplier information attached to only 80 % of the components can cause
puzzlement in the purchasing department - what about the 20 % of
components that are already sourced, but do not have a supplier according to the new system?
- A newly created item missing 20 % of its attribute
values can cause headaches downstream in Production
Obviously, in these cases, 100 % correctness should be the target. While this
might be a challenging goal, the cost of poor quality can become very
high when wrong and/or incomplete data impacts on production processes
and customer satisfaction.
Another example where the 80-20 rule does not work well on (product) data
is when configuration tables in a PDM system are used to manage rules
such as: If customer X belongs to market segment A, then select option 1. Such
configuration tables should be fully functional and tested before use,
as customer X won't be too happy receiving a product with the wrong options.
PDM systems (with their appealing workflows, authorization and access management
mechanisms) and configuration management, combined with overly enthusiastic
business requirements, can easily become very complex.
For such complex PDM systems, the number of (theoretically) required test
cases can be so high that 80-20 needs to be applied on the testing effort
itself. We get into confidence and probability analysis:
- If we put X amount of effort into testing, then we have a Y % chance of
missing an error that will cost our business Z dollars, or
- If we put P amount of effort into additional testing, we can
reduce the chance of expensive errors by Q %
I wonder how many of the companies that went through the pain of putting
a PDM system into use realized at the beginning of their project that such
high-effort testing and validation activities reoccur when changes and
upgrades are made to their systems.
So, because product data is at the core of PLM, be careful when applying Pareto's
Principle. It might cost you 80 % of your project budget to fix the 20 %
of product data quality issues that occur after the project is "finished".
Menno Huijben is a senior advisor at Sofigate Oy. Sofigate is dedicated to improve
the performance of Information Management in organizations, hands-on and down-to-earth.
Menno can be contacted at an address of the type
firstname.secondname@sofigate.com
* The Evolution of PLM Theory *
By John Stark
An attendee at the recent PLM Summit asked about the evolution of PLM theory. The current theory states
that:
- PLM is an essential business activity. PLM is the business
activity of managing a company's products all the way across their lifecycles. PLM manages the
product from cradle to grave, from sunrise to sunset.
There are several phases in a product's lifecycle, e.g., the five phases of imagination,
definition, realisation, use/support/maintenance and retirement/disposal/recycling.
- PLM is holistic.
PLM addresses nine components: the product; data and documents; processes; applications;
methods; facilities and equipment; people; organisational structure; metrics.
It doesn't address just one resource,
and improve use of that resource while reducing the effectiveness of other resources.
- The two axes of components and lifecycle phases
lead to a 5 * 9 PLM Grid with 45 cells.
- PLM is a "joined-up" paradigm. It joins up many things that
were previously separate. In the past, products were to a certain extent managed across the
lifecycle, so a lot of the components needed for PLM already exist. However, there are gaps,
contradictory versions of the same data, information sinks, silos, non-functioning networks,
duplicate processes, ineffective fixes, etc., that result in reduced revenues and higher costs.
- PLM puts the focus on the product. Customers buy products,
corporate revenues result from sales of products.
- PLM gets products under control across the lifecycle, reducing
product-related risk and providing visibility about the product across the lifecycle.
- PLM has business objectives. It aimes to maximise the value
of the product portfolio, increase product revenues and reduce product-related costs.
- There are many reasons why PLM is needed, such as needs to:
develop and support products globally; develop products faster; comply with
regulations, etc.
- With PLM, the CEO can give responsibility for all the
products to one person - the Chief Product Officer (CPO).
- To make PLM happen, the CEO should launch a PLM Initiative.
This can follow the Ten-Step Approach to PLM:
PLM Status Review and Data Gathering; Executive PLM Education and Awareness;
Best Practice Positioning; PLM Concept Generation and Analysis;
PLM Roadmap and Plan Generation; Business Benefits and Business Case Development;
ROI Calculation; Management Report Preparation; Executive Presentation; Executive Decision Support
* Global PLM 2007 Survey *
The Global PLM 2007 Survey is a survey of the status of PLM practice
and thought in manufacturing and engineering companies throughout the world in 2007.
The current number of responses is about 90% of the target number.
As a result, we are extending until August 2007 the timeframe of Part 1 of the survey.
Participant feedback is that the questionnaire for the first part of the
survey takes about 10 minutes to answer.
More details here.
|
