Today I have been...
Reading about Statistical Process Control.
SPC is designed for solving problems related to "process variation".
For example, a process maybe unable to meet quality or accuracy standards desired or required. The concepts of Technical criticality (where a process fails to meet specifications resulting in defective products/services) and Statistical criticality (where a process is unstable or incapable of meeting its performance specifications).
SPC is a big topic in T889 and there is a question in TMA02 on it. Mason and Antony provide contextual information in S11.5 B3.
Tuesday, 31 January 2012
Today I have been...
(continued)
Reading about Y2X. It builds on the I/O of SIPOC charts and can be done ahead of other improvement methods and techniques. A matrix with outputs on the Y, and inputs on the X it shows how I's influence O's.
The next is Taguchi methods which are used for optimising product and process, and pre-manufacture design tests.
It is very heavy on statistical methods (more than are covered in T889) and I am concerned that so many of these methods seem to focus on production process improvement, rather than service or even "intangibles" such as software. See S10.5 in B3, there is much material.
SMED is next up, and is closely associated with lean but can be used on its own to help solve throughput problems, according to the text. Often, it is used in batch manufacture to help reduce the inter-batch time gap (the name comes from a process like this - single minute exchange of dies)
Process redesign is covered next and this definitely can be used for services. The starting point for this activity is to use something along the lines of activity sequence diagrams, flow-block diagrams, flow process diagrams or spaghetti charts (all covered in t889) to model the existing process. It's also important to understand the customer interface "touch points" for service processes. The text focuses on throughput time analysis and bottleneck analysis. Slack (2007) lays out principles of an "improvement-oriented approach" to process redesign. An interesting point is that "an hour saved at a non-bottleneck is a mirage. Non-bottlenecks have spare capacity anyway. Why bother making them even less utilised?" and another is that something is only utilised if it contributes to the entire process, in other words if it spends 100% of its time on value-add activity.
Slack's points on page 99 & 100 of the block seem to be extremely pertinent and useful.
(continued)
Reading about Y2X. It builds on the I/O of SIPOC charts and can be done ahead of other improvement methods and techniques. A matrix with outputs on the Y, and inputs on the X it shows how I's influence O's.
The next is Taguchi methods which are used for optimising product and process, and pre-manufacture design tests.
It is very heavy on statistical methods (more than are covered in T889) and I am concerned that so many of these methods seem to focus on production process improvement, rather than service or even "intangibles" such as software. See S10.5 in B3, there is much material.
SMED is next up, and is closely associated with lean but can be used on its own to help solve throughput problems, according to the text. Often, it is used in batch manufacture to help reduce the inter-batch time gap (the name comes from a process like this - single minute exchange of dies)
Process redesign is covered next and this definitely can be used for services. The starting point for this activity is to use something along the lines of activity sequence diagrams, flow-block diagrams, flow process diagrams or spaghetti charts (all covered in t889) to model the existing process. It's also important to understand the customer interface "touch points" for service processes. The text focuses on throughput time analysis and bottleneck analysis. Slack (2007) lays out principles of an "improvement-oriented approach" to process redesign. An interesting point is that "an hour saved at a non-bottleneck is a mirage. Non-bottlenecks have spare capacity anyway. Why bother making them even less utilised?" and another is that something is only utilised if it contributes to the entire process, in other words if it spends 100% of its time on value-add activity.
Slack's points on page 99 & 100 of the block seem to be extremely pertinent and useful.
Monday, 30 January 2012
A group of techniques
Today I have been...
Looking at the techniques block of T889.
It offers various problem analysis tools....
Pareto analysis
Cause-and-effect diagrams
Stratification
Tally cards
Histograms
Scatterdiagrams
Shewhart control charts
It calls these the 7 old tools.
It then discusses 7 new tools
Affinity diagrams
Relational diagrams
Tree diagrams
Matrix diagrams
Program decision process charts
Arrowdiagrams
Matrix diagram analysis
It then covers questioning techniques - is/is not, 5 whys, 5 w+h, and more specific question sets.
Next, diagramming techniques are covered
Systems diagrams
Input-output diagrams
Systems maps
Influence diagrams
Richpictures
Other diagramming techniques
Activity sequence diagrams
Flow-block diagrams
Flow-process diagrams
Spaghetti charts
SIPOC charts
Multiple-cause diagrams
Force field analysis
Cognitive mapping
There is some discussion then around FMEA and Fault Tree Analyses.
There is a section on Creativity and idea generation which is similar to block 2 of B822
This covers Brainstorming, Brainwriting, Nominal group techniques, SCAMPER,Creativeproblem solving and finally De Bono's Six hats and lateral thinking.
There is information on techniques used for examining context such as environmental scanning, stakeholder analysis, SWOT, benchmarking and gap analysis. Berry et al is cited regarding gap analysis.
It then covers design techniques...
The first of these is the interestingly-named poka-yoka (avoiding inadvertant errors) - this could maybe used in conjunction with FMEA.It also covers quality function deployment (the house of quality - a conceptual framework that can be used to guide teams through the transformation process that converts customer requirements into successful products).
B3P78 shows this concept and the elements of the QFD process
1. Customer Requirements
2. Importance Weighting
3. Design requirements
4. Relationship matrix
5. Correlation matrix
6 & 7 Competitive Assessments or benchmarking
8. Objective measures
The output of the house of quality is a set of objective measures that relate to the substitute quality characteristics. The next "house" is parts deployment, which takes these design requirements and produces parts characteristics. The Process Planning house then takes the parts characteristics and produces key manufacturing operations. The final Production Planning house takes the key manufacturing operations and produces production requirements. There are four houses/phases in all. I wonder about the applicability of QFD to software solution design, particularly where there is a desire to be agile.
The section then goes on to discuss TRIZ. It is argued that there are few real-life examples of finished products where it has been used as part of the design process. The author, despite this, thinks it has two valuable and insightful concepts, "ideality" and "contradictions". Martin G Moehrle's paper on TRIZ "What is TRIZ? From Conceptual Basics to a Framework for Research" is given as an source of further information.
The conceptual approach of TRIZ is
Concrete Problem -> Abstract Problem -> Abstract Solution -> Concrete Solution
It provides a framework for moving through these steps to solve product and service problems.
(to be continued)
Why?
So What?
How will I use it?
Looking at the techniques block of T889.
It offers various problem analysis tools....
Pareto analysis
Cause-and-effect diagrams
Stratification
Tally cards
Histograms
Scatterdiagrams
Shewhart control charts
It calls these the 7 old tools.
It then discusses 7 new tools
Affinity diagrams
Relational diagrams
Tree diagrams
Matrix diagrams
Program decision process charts
Arrowdiagrams
Matrix diagram analysis
It then covers questioning techniques - is/is not, 5 whys, 5 w+h, and more specific question sets.
Next, diagramming techniques are covered
Systems diagrams
Input-output diagrams
Systems maps
Influence diagrams
Richpictures
Other diagramming techniques
Activity sequence diagrams
Flow-block diagrams
Flow-process diagrams
Spaghetti charts
SIPOC charts
Multiple-cause diagrams
Force field analysis
Cognitive mapping
There is some discussion then around FMEA and Fault Tree Analyses.
There is a section on Creativity and idea generation which is similar to block 2 of B822
This covers Brainstorming, Brainwriting, Nominal group techniques, SCAMPER,Creativeproblem solving and finally De Bono's Six hats and lateral thinking.
There is information on techniques used for examining context such as environmental scanning, stakeholder analysis, SWOT, benchmarking and gap analysis. Berry et al is cited regarding gap analysis.
It then covers design techniques...
The first of these is the interestingly-named poka-yoka (avoiding inadvertant errors) - this could maybe used in conjunction with FMEA.It also covers quality function deployment (the house of quality - a conceptual framework that can be used to guide teams through the transformation process that converts customer requirements into successful products).
B3P78 shows this concept and the elements of the QFD process
1. Customer Requirements
2. Importance Weighting
3. Design requirements
4. Relationship matrix
5. Correlation matrix
6 & 7 Competitive Assessments or benchmarking
8. Objective measures
The output of the house of quality is a set of objective measures that relate to the substitute quality characteristics. The next "house" is parts deployment, which takes these design requirements and produces parts characteristics. The Process Planning house then takes the parts characteristics and produces key manufacturing operations. The final Production Planning house takes the key manufacturing operations and produces production requirements. There are four houses/phases in all. I wonder about the applicability of QFD to software solution design, particularly where there is a desire to be agile.
The section then goes on to discuss TRIZ. It is argued that there are few real-life examples of finished products where it has been used as part of the design process. The author, despite this, thinks it has two valuable and insightful concepts, "ideality" and "contradictions". Martin G Moehrle's paper on TRIZ "What is TRIZ? From Conceptual Basics to a Framework for Research" is given as an source of further information.
The conceptual approach of TRIZ is
Concrete Problem -> Abstract Problem -> Abstract Solution -> Concrete Solution
It provides a framework for moving through these steps to solve product and service problems.
(to be continued)
Why?
So What?
How will I use it?
Sunday, 20 November 2011
Today I have been...
Reading the special issue of Academy of Management Learning and Education 2005, vol 4 no 2, where there are 7 reviews of Henry Mintzberg's criticism and views on (North American) MBA education in his book Managers not MBAs.
Why?
Activity 1.1, B830 Issues and Approaches - part 1, Integrating Practice, Learning and Theory.
So What?
Mintzberg's view is that (at the time of writing) the majority of MBA students were too young to do an MBA course. His view was they should have engaged some management experience first, and in his response to the reviews he asks three questions:-
The main thing I have noticed is my MBA studies have given me almost a different "lens" to view my work with. The models and frameworks and techniques taught have become part of a toolset, and I certainly feel better able to interact with senior staff within the organisation at their level.
I have also hugely valued the networking that doing this study has enabled. I have met some others who have studied the same qualifications and modules within my workplace and outside it, and have enjoyed working with others from a cross-section of industries at study groups, tutorials and residential schools.
The IMPM approach as described by Mintzberg seems to be trying to address his concerns. The main attraction is the notion of it being practice-based. There is a difference about training someone to talk about good practice and training them to practice good practice.
How will I use it?
I hope the remainder of my study will help develop the three styles of management in me - calculating, heroic and engaging - and particularly engaging.
Reading the special issue of Academy of Management Learning and Education 2005, vol 4 no 2, where there are 7 reviews of Henry Mintzberg's criticism and views on (North American) MBA education in his book Managers not MBAs.
Why?
Activity 1.1, B830 Issues and Approaches - part 1, Integrating Practice, Learning and Theory.
So What?
Mintzberg's view is that (at the time of writing) the majority of MBA students were too young to do an MBA course. His view was they should have engaged some management experience first, and in his response to the reviews he asks three questions:-
- Does the MBA class consist of practicing managers? Managers cannot be created in a classroom, trying to do so just creates hubris.
- Do these managers stay on the job while studying? Ripping them out of their jobs to study full-time foregoes the opportunity to develop them in the context of their practice.
- If these managers are on the job, does the program(me) focus on them learning from their own experience? Not just other people's experience, as in cases and theory, but their own experience. The former are fine but limited; the latter is key. And not just at coffee breaks and mealtimes, but centrally, concertedly, in the classroom - for example half its time devoted to managers sharing experiences among themselves, on their agendas.
The main thing I have noticed is my MBA studies have given me almost a different "lens" to view my work with. The models and frameworks and techniques taught have become part of a toolset, and I certainly feel better able to interact with senior staff within the organisation at their level.
I have also hugely valued the networking that doing this study has enabled. I have met some others who have studied the same qualifications and modules within my workplace and outside it, and have enjoyed working with others from a cross-section of industries at study groups, tutorials and residential schools.
The IMPM approach as described by Mintzberg seems to be trying to address his concerns. The main attraction is the notion of it being practice-based. There is a difference about training someone to talk about good practice and training them to practice good practice.
How will I use it?
I hope the remainder of my study will help develop the three styles of management in me - calculating, heroic and engaging - and particularly engaging.
Wednesday, 9 November 2011
Metaphors for problem solving
Three generic metaphors describe three problem solving/improvement methods - a "learning cycle", a "journey", a "search". See also B822 - problem solving as cultivation, search etc.
Best known learning cycle is "PDCA" - plan-do-check-act. This is similar to Kolb's learning cycle and often called the Deming wheel or the Shewart cycle. Diagram b1 p41
Oakland and Marosszeky call this the DRIVE modle for continuous improvement - Define, Review, Investigate, Verify, Execute.
The cycle delivers continuous improvement. It is a cycle because it does not end.
The journey metaphor allows you to think of PS&I as a way to move from where you are now to where you want to be. As with any journey differing routes are often possible, and each will have its advantages and disadvantages. See figure 1.5 p44
Stage 1 - where are you now
Stage 2 - where would you like to get to and what is stopping you
Stage 3 - how do you know when you have arrived
Stage 4 - how could you get there
Stage 5 - how can you tell the outcomes
Stage 6 - what are the outcomes
Stage 7 - choose the best route
Stage 8 - OK, go ahead.
Stage 3 is where you establish how to measure performance, so you need to make sure you state what those measures are, what the targets are, what the timescales are and what are the acceptable deviations from those targets (think SMART objectives). All objectives identified during stage 2 need to be included.
For example, three measures that could be use to manage the performance of a purchasing department include money saved, purchase order processing time, and customer satisfaction (accuracy) (time, cost, quality).
For stage 7 - the 5W+H can be useful - what, why, where, when, how, who?
Problem solving as search is similar to the journey metaphor - see B1 Page 46.
Best known learning cycle is "PDCA" - plan-do-check-act. This is similar to Kolb's learning cycle and often called the Deming wheel or the Shewart cycle. Diagram b1 p41
Oakland and Marosszeky call this the DRIVE modle for continuous improvement - Define, Review, Investigate, Verify, Execute.
The cycle delivers continuous improvement. It is a cycle because it does not end.
The journey metaphor allows you to think of PS&I as a way to move from where you are now to where you want to be. As with any journey differing routes are often possible, and each will have its advantages and disadvantages. See figure 1.5 p44
Stage 1 - where are you now
Stage 2 - where would you like to get to and what is stopping you
Stage 3 - how do you know when you have arrived
Stage 4 - how could you get there
Stage 5 - how can you tell the outcomes
Stage 6 - what are the outcomes
Stage 7 - choose the best route
Stage 8 - OK, go ahead.
Stage 3 is where you establish how to measure performance, so you need to make sure you state what those measures are, what the targets are, what the timescales are and what are the acceptable deviations from those targets (think SMART objectives). All objectives identified during stage 2 need to be included.
For example, three measures that could be use to manage the performance of a purchasing department include money saved, purchase order processing time, and customer satisfaction (accuracy) (time, cost, quality).
For stage 7 - the 5W+H can be useful - what, why, where, when, how, who?
Problem solving as search is similar to the journey metaphor - see B1 Page 46.
Tuesday, 1 November 2011
Approaches to problem solving and improvement
Reminder: Problem solving and improvement are best done using 'systematic rational thought supported by relevant information' (Gruenig and Kuehn, 2005, p8)
How do you make sure you are doing this?? Use an approach or method.
This forces the user to ensure that bits that might conveniently be omitted are confronted and examined. It also provides a way for someone looking to verify the results to ensure that the problem was tackled in a sensible way, allowing confidence in the solution.
Approaches can be reductionist (designed to solve science problems - see also Portal and Portal 2) or holistic.
Reductionist approaches, according to waddington 1977 require
1. devising alternative hypotheses
2. devising a crucial experiment .... [to] exclude one or more of the hypotheses
3. carry out the experiment...and recycle the procedure making sybhypotheses or sequential hypotheses to define the possibilities that remain
^^^ the "method of strong inference"
A holistic approach considers the reactions between components to be just as important as the components themselves.
Block 1 p38 table 1.2 shows a comparison between holistic and reductionist approaches.
An alternative discriminator is to consider whether approaches are heuristic or analytic. Heuristics are rules of thumb, strategies, tricks, simplifications etc. A way to limit search for solutions in large problem spaces. A useful heuristic "offers solutions which are good enough most of the time". Examples are means-end analysis and planning.
Analytic problem solving uses techniques to conduct rigorous analysis. There may be limitations as to when they can be applied.
Solutions based on analytic techniques can be assessed by criteria in four areas
1. content (what will be the outcome)
2. level (a kind of performance indicator)
3. how long the solution will be valid for
4. its scope
Diagram 1.2 p40 block 1
Pick two out of three: cheap, fast or effective. You can't have all - related to time/quality/cost.
How do you make sure you are doing this?? Use an approach or method.
This forces the user to ensure that bits that might conveniently be omitted are confronted and examined. It also provides a way for someone looking to verify the results to ensure that the problem was tackled in a sensible way, allowing confidence in the solution.
Approaches can be reductionist (designed to solve science problems - see also Portal and Portal 2) or holistic.
Reductionist approaches, according to waddington 1977 require
1. devising alternative hypotheses
2. devising a crucial experiment .... [to] exclude one or more of the hypotheses
3. carry out the experiment...and recycle the procedure making sybhypotheses or sequential hypotheses to define the possibilities that remain
^^^ the "method of strong inference"
A holistic approach considers the reactions between components to be just as important as the components themselves.
Block 1 p38 table 1.2 shows a comparison between holistic and reductionist approaches.
An alternative discriminator is to consider whether approaches are heuristic or analytic. Heuristics are rules of thumb, strategies, tricks, simplifications etc. A way to limit search for solutions in large problem spaces. A useful heuristic "offers solutions which are good enough most of the time". Examples are means-end analysis and planning.
Analytic problem solving uses techniques to conduct rigorous analysis. There may be limitations as to when they can be applied.
Solutions based on analytic techniques can be assessed by criteria in four areas
1. content (what will be the outcome)
2. level (a kind of performance indicator)
3. how long the solution will be valid for
4. its scope
Diagram 1.2 p40 block 1
Pick two out of three: cheap, fast or effective. You can't have all - related to time/quality/cost.
Monday, 31 October 2011
Putting your improvement into boxes (categorising it!)
There are various categories of improvement, perhaps the best known being "continuous improvement" (CI). CI is linked to modern quality management.
Quality management's three prongs:-
quality control
quality improvement
quality planning
Feigenbaum's four steps of quality control
Setting Quality standards
Appraising conformance to those standards
Acting when the standards are exceeded
Planning for improvements in the standards
Quality Gurus exist. Deming is one of them. Deming's 14 point plan for the achievement of Total Quality Management (TQM) focuses heavily on improvement (b1p33)
Continuous improvement is often referred to by the Japanese word Kaizen. (ref B822 again).
Examples of kaizen
Quality Circles
Suggestion Schemes
Other tools include benchmarking, traffic light schemes, kaikaku or breakthrough (radical improvement) and kairoy (improvement achieved through innovation and investment in new plant or systems).
Juran defined breakthrough as the following sequence for solving chronic problems:-
1. Convince others that a breakthrough is needed - that a change in quality level is desirable and feasible
2. Identify the vital few projects - determine which quality problem areas are most important
3. Organise for breakthrough in knowledge - define the organisational mechanisms for obtaining missing knowledge
4. Conduct the analysis - collect and analyse the facts that are required and recomment the action needed
5. Determine the effect of proposed changes on the people involved and find ways to overcome the resistance to change.
6. Take action to institute the changes
7. Institute controls to hold the new level
There are two types of innovation - incremental and radical (b822 again - adaptor/innovator).
Leach et al performed a study to examine the attributes of more successful major innovative companies. They are in box 1.4 p36 b1.
Quality management's three prongs:-
quality control
quality improvement
quality planning
Feigenbaum's four steps of quality control
Setting Quality standards
Appraising conformance to those standards
Acting when the standards are exceeded
Planning for improvements in the standards
Quality Gurus exist. Deming is one of them. Deming's 14 point plan for the achievement of Total Quality Management (TQM) focuses heavily on improvement (b1p33)
Continuous improvement is often referred to by the Japanese word Kaizen. (ref B822 again).
Examples of kaizen
Quality Circles
Suggestion Schemes
Other tools include benchmarking, traffic light schemes, kaikaku or breakthrough (radical improvement) and kairoy (improvement achieved through innovation and investment in new plant or systems).
Juran defined breakthrough as the following sequence for solving chronic problems:-
1. Convince others that a breakthrough is needed - that a change in quality level is desirable and feasible
2. Identify the vital few projects - determine which quality problem areas are most important
3. Organise for breakthrough in knowledge - define the organisational mechanisms for obtaining missing knowledge
4. Conduct the analysis - collect and analyse the facts that are required and recomment the action needed
5. Determine the effect of proposed changes on the people involved and find ways to overcome the resistance to change.
6. Take action to institute the changes
7. Institute controls to hold the new level
There are two types of innovation - incremental and radical (b822 again - adaptor/innovator).
Leach et al performed a study to examine the attributes of more successful major innovative companies. They are in box 1.4 p36 b1.
Subscribe to:
Posts (Atom)