Body of Knowledge for Classical TRIZ
Editor | On 01, Oct 2007
By D. Daniel Sheu
This paper provides an overview of the classical TRIZ (Theory of Inventive Problem Solving) body of knowledge. The author classified two ways of using TRIZ tools: by “contents” of generic solutions and by following specific problem analyzing/solving “processes.” Content-oriented TRIZ tools may be industry or application specific. Process-oriented TRIZ tools are, in general, applicable to all industries or applications. The functions of major tools are explained briefly and the relationships among the various TRIZ tools are presented. Contributions of this research include: 1) identifying TRIZ working principles, 2) establishing a map of the TRIZ body of knowledge (including the identification of content- vs. process-oriented solution tracks, explaining why some content-oriented tools are not fully applicable to different industries) and 3) identifying the relationship among the tools of classical TRIZ.
TRIZ, body of knowledge, classical TRIZ, systematic innovation
For the purpose of better training, this paper organizes the body of knowledge for classical TRIZ (Theory of Inventive Problem Solving). The essences of the main tools are briefly explained, along with their working principles and division info categories. The paper mainly focuses on classical TRIZ – modern extensions are not included.
TRIZ: Overview and Working Principles
TRIZ Fundamental Beliefs
TRIZ consists of a number of tools, forming a logical sequence of process guided by a set of fundamental philosophies as shown in Figure 1.2The fundamental beliefs of TRIZ can be classified into four categories:
- Problem solutions: We believe that if we analyze the problem into more fundamental level, in a great majority of cases, someone somewhere already (intelligently) solved a problem similar to ours. Therefore, we should take advantage of the prior wisdom for problem solutions.
- Problem solving processes: Like solutions, problem solving processes also have patterns that can be followed. Someone, somewhere, has used certain processes to solve similar problems before. It is to our advantage to follow the processes in order to solve our problem.
- Trends: The technical evolutions have trends. If we observe the trends, the future stages of the trends applicable to our problem provides guidance on how to solve our current problem.
- Goals: The ideal final result (IFR) provides clear direction for any and all goal(s).
These four fundamental beliefs provide systematic approaches to solving problems.
TRIZ Working Principles
The author has identified several underlying phenomena that make TRIZ work:
Classification has been one of the most effective methods that humans use to understand and solve problems. When humans try to understand things, they begin by classifying them. Items in the same class bear certain similar attributes and we use certain class of tools to deal with them – there is no panacea. In TRIZ, problems are classified into a number of patterns based on certain classification perspectives and represented in certain models. If we classify our subject problem into an existing class, we can use the existing corresponding solutions to solve said problem. The most common classification perspectives classical TRIZ used are: configurations on parameters relationships, which led to the contradiction matrix (CM – modelling of the problem) and the 40 inventive principles (IP – solutions), and configurations on substance/field relationships, which led to the substance-field analysis (SFA – modelling of problems) and 76 inventive standards.
In a sense, the contradiction matrix and substance-field analysis were developed based on the same principle of classification. The only difference is the way of classifying and modelling the problems. The CM uses conflicting parameters to classify problems while the SFA uses relationships among interacting substances and their fields to classify and model problems, thus leading to two different sets of problem solutions. The author infers that if we can find some other generic perspective to classify problems, we may be able to develop new representations of standard solutions to use.
2: Commonality and Reduction of Number of Classes at Fundamental Level
There are innumerable types of objects in the world – tens of millions at the least. However, if we divide the objects into their parts and further down to the fundamental physical/chemical level, the types of elements are now greatly reduced. When we divide the objects into their elements, there are only about one hundred elements; those that exist are quite organized. Consider this in another way – there are only 26 letters in the English alphabet and about a hundred or so root patterns. Based on these relatively very few elemental character strings, we can produce essentially infinite number of words and sentences to use. This tells us that seemingly different classes of problems viewed at the human interface level may become the same in the fundamental physical/chemical level. This allows us to transform problems from the human level to the fundamental level and find that 99.99 percent of our problems at the fundamental level are the same as some other problems previously solved.
This commonality at the fundamental level enables us to leverage the intelligent solutions found before. This is the heart of TRIZ. Solving problems at the fundamental level has the following benefits: 1) it is simpler and easier because it involves only a small number of elements and attributes, 2) it is likely solved already by someone before, so we can leverage the prior wisdom and 3) it allows us to quickly focus on the key area without affecting most other areas.
3: Contents Vs. Processes
There are patterns of TRIZ processes as well as contents. Content-oriented tools provide specific classes of solutions with respect to a given specific class of problems. Process-oriented tools do not provide solutions directly. Instead, they provide processes along which one can follow to develop solutions. The main doubts/complaints on the applicability of TRIZ tools are centered on some of the content-oriented tools such as the contradiction matrix and inventive principles. In recent mechanical inventions, there is a low 48 percent applicability of the CM and IP.1 A thesis study reported that the classical CM and IP only showed a 45 percent success rate of applicability for 22 patents randomly selected granted in 2007 on chemical mechanical processing machines (CMP) in the semi-conductor industry. A reproduced CM and IP based on more recent patents showed that CMP patents had a 77 percent success rate of applicability for the same test set of CMP patents.3,4 Even though the sample size is small, a statistical test proved that the difference in applicability is statistically significant:
- The number of cases needed to develop a CM and IPs for a particular industry could be much less than that in the classical CM and IPs, if the patent cases used are from the same industry/application of interest.5,6 This is because the large number of patents collected from drastically different industries add little applicability to explain the heterogeneous fundamental physics/chemistry of test cases. On the other hand, CM/IP developed from patents of like industries/applications can explain better due to similar underlying fundamental physics/chemistry.3,4
Body of Knowledge for Classical TRIZ
There is a path to take with the body of knowledge for classical TRIZ and how to use it to solve problems. (See Figure 2.)We start with a specific problem to solve. The specific problem is analyzed using problem exploration, understanding and definition tools so that we can assess the best means of attack. This process helps look at the alternatives of a broader or narrower problem, and the choices with which we can decide mini-problem to solve.
Once the “right” problem is selected and well understood, there are two generic paths to take for solving the problem – using content-oriented tools and/or process-oriented tools. We may test both paths to get the most out of TRIZ. Tools such as the contradiction matrix, inventive principles, Su-Field analysis, etc., are content-oriented because the solutions for classes of problems are already available. In the process-oriented path, the tools provide the processes to follow so that we can develop the solutions ourselves – instead of giving us the solution triggers directly as do content-oriented tools.
On the content-oriented path, we classify and model the problem into a class of problem; the corresponding solutions can be used immediately. By doing so, we are not doing much innovative work although the solution itself may be innovative. On the process-oriented path, although we do not have solutions handy, the provocative thinking processes required by the tools put us in an environment that is conducive to innovation – enabling us to draw creative solutions by ourselves. Unlike the contents-oriented tools that may be industry or process/product specific, the process-oriented tools are, in general, applicable across all industries/products/processes. In addition, the solution obtained by the process-oriented tools will be, in general, specific solutions to the specific problem. It is important to note that the solutions obtained from the content-oriented tools are generic “trigger” solutions. Some additional thinking may be needed to convert from the generic solutions into specific solutions. It is usually more complex and time-consuming to take the process-oriented path rather than the content-oriented.
Once the specific solutions are found, a systematic solution evaluation is needed to determine feasibility and desirability of the solutions. The solution(s) that pass the evaluation and selection processes are then implemented by project planning, execution, etc. The final results are then checked for meeting the expected benefits and follow-up monitoring/control. Hopefully, the problem is now solved satisfactorily.
Relationships Among the Tools
Problem Exploration and Definition
The problem explorer and definition tools include benefit analysis, problem hierarchy analysis, resource analysis, constraints analysis, etc. These tools help us to make sure that:
- The project we are working on is worthy of doing (benefit analysis),
- We are working on the right/most appropriate problem to attack (problem hierarchy analysis),
- We fully understand the constraints of the problem,
- We examined the available resources for possible usage, and
- We know the location of our key sore-point.
The problem exploration and definition tools help us focus on the right problem at the right position, with the problem contents and boundary well-defined and positioned ready for further analysis.
Function Attribute Analysis and Forward Thinking Tools
Function attribute analysis (FAA) takes the problem as the system of analysis, breaks it down into components, and identifies the relationships among the components and the attributes of those components. FAA helps us further understand the system in detail and formulate it into a format ready for the following TRIZ solving tools:
- Knowledge effect database: When FAA identifies the needed function(s) to achieve, or attribute to manipulate, system improvement, the knowledge effect database (KE/DB) can be used to locate the means to achieve those functions and attribute altering.
- Contradiction matrix and inventive principles: When FAA identifies a component that creates both negative function and positive function, we can review the involved attributes (parameters) to locate the contradiction between the attributes. If there are two different conflicting attributes, we can use the CM/IP to resolve the problem.
- Physical contradictions and separation principles: If the conflicting attributes are the same or there is a controlling attribute to control the two conflicting attributes, the physical contradiction is located and the separation principles used to resolve the conflict.
- Trimming or increasing functions: Once all the functions of the system are identified by FAA, non-primary functions are natural targets for trimming or increase to improve system ideality. Depending on its S-curve stage, trimming or increasing functions can improve the system. When the system is at the earlier part of the S-curve, the system can be effectively improved by increasing functions. If the system is at a later stage of the S-curve, trimming and the ideal final result (IFR) can be used to improve the system; the IFR can be used as the guide for trimming.
- Patent evasion and road mapping: We may use FAA to analyze a system for patent-related reasons. Once the components/functions already patented by others are identified, we know what functions/components we may claim for our own patents. We can try substituting alternative components or functions to evade existing patents. By looking at all involved components/functions, we can plan our patent roadmap to strategically build effective patent protection.
- Mini-problem identification in ARIZ: The mini-problem used in the algorithm for inventive problem solving (ARIZ) can be easily identified from the results of FAA. It usually starts from a negative function on the FAA where a conflicting pair of attributes/functions can be located – this relates FAA and ARIZ. FAA is a fundamental part of the overall ARIZ process. All the tools used here are natural steps forward from FAA – associated with forward thinking.
Black type denotes tool names; red type describes the relationships; blue lines and type indicate the comparisons.
S-Curve, Evolutionary Trends and Leap-frog Thinking
Technology S-curves usually use some sort of ideality measure on the vertical axis and some form of timing measure on the horizontal side, thus relating S-curve and ideality. Each stage of evolutionary trends is an S-curve by itself. When we use evolutionary trends as solution triggers for our improvement ideas, we are leap-frogging forward on system improvements, as the ideas hinted by the evolutionary stages signify disruptive jumps. We use the best of existing technology stages to help us jump forward one or several stages. All the technical evolutionary stages develop toward increasing ideality at the system level and, also, in normal cases, at the components level – this relates trends and ideality.2
Three types of improvement thinking are contrasted here. There is no clear-cut answer for which way is better – examining all three ways of thinking can lead to improvements.
- Leap-frog forward thinking: As in the case of using evolutionary trends for improvements
- Backward thinking: As in the case of using the IFR to guide our ideas for improvement –start with the IFR and gradually step backward until a doable issue is reached
- Step forward thinking: As developed step-by-step from problem definition to FAA and to the various conflict resolution tools
Ideality, IFR and Backward Thinking
Ideality is defined as: perceived benefit/(cost + harm). It is a measure of ultimate performance. The achievement of highest ideality is the IFR. The IFR concept allows for powerful backward thinking to achieve the best possible solution – fast. Trimming can look to the IFR for guidance and the S-curve can use ideality as its performance measure.
The use of resources encourages us to minimize the use of resources and use all free resources as much as possible. This contributes to the minimization of cost on the ideality equation for improvements. Each previous stage on the evolutionary trend can be used as the resource for the next stage. This relates the IFR, resources and trends of evolution.
Tools for Changing Perspectives
Psychological inertia (PI) constitutes a barrier that denies us from seeing things with different perspectives. The benefits of thinking from different perspectives allow us to solve problems in a more effective and efficient way – a kind of paradigm shift. Some problems may be difficult to see or solve in one paradigm, yet easily seen or solved in a different one. Nine windows allows us to imagine any subject tool being applied across past-present-future and component-system-super-system perspectives. The 9 windows tool of PI can be applicable to problem definition tools, FAA, S-curve, evolutionary trends, resources, ideality, etc. Even with the separation principles, we can consider transferring from component to system or from one system to another system – also variants of 9 windows thinking. The 9 windows view is an expansion of our regular perspective. Almost its opposite, the smart-little-people (SLP) tool “shrinks” our perspective to a micro-level.
ARIZ is the integration of all the major TRIZ tools working in a systematic procedure.
The Taiwan TRIZ Association’s Curriculum
The TRIZ body of knowledge was divided into a series of eight short courses, plus one software course. The main concern is to allow industry people to learn the courses in piecemeal – accommodating their work – while presenting them in a comprehensive yet easy-to-learn manner.
The courses are organized in three groups:
Group A: Core Courses
- Introduction to TRIZ Theory & Practices: serves as a general introduction to TRIZ; includes the origin of TRIZ, how it works, what each tool does in a nutshell and the relationships among the tools; levels of innovation (6 hrs)
- Contradiction Matrix & Invention Principles: covers the details of technical and physical contradictions, the contradiction matrix and inventive principles, and the separation principles(6 hrs)
- Su-Field & Inventive Standards (6 hrs)
- Trimming & Simplicity (6 hrs)
- Ideality & Trends of Technical Evolution: includes the S-curve(6 hrs)
- Algorithm for Inventive Problem Solution (ARIZ) (12 hrs)
Group B: Derivative Courses
- Patents and TRIZ: includes patent strategy and how to use FAA for patent evasion and roadmapping(6 hrs)
- Green Design and TRIZ (6 hrs)
Group C: TRIZ Software Briefing
- CREAX Software (3 hrs)
- I-TRIZ Software (3 hrs)
- Goldfire software (to be added – 6 hours)
The knowledge map for classical TRIZ allows for an interpretation of its fundamental working principlesand provides the roadmap forTRIZ study established by the Taiwan TRIZ Association. By clarifying the difference between content-oriented and process-oriented tools, it becomes clear that problems must be solved with respect to their relevant industries and families.
The author acknowledges the assistance of Mr. K. Ko and Y. Wang with this paper’s dictation.
- Mann, D. L., “Assessing the Accuracy of the Contradiction Matrix for Recent Mechanical Inventions,” The TRIZ Journal, February 2002.
- Mann, D. L., Hands-on Systematic Innovation, CREAX Press, ISBN 90-77071-02-4.
- Sheu, D. Daniel and Pang-Yen Yu, “Invention Principles and Contradiction Matrix for Semiconductor Manufacturing Industry: Chemical Mechanical Polishing”, The 2nd Annual Meeting & Conference of the Taiwan TRIZ Association, December 15, 2007, National Tsing Hua University, Hsinchu, Taiwan (in Chinese).
- Yu, Pang-Yen, “Invention Principles and Contradiction Matrix for Semiconductor Manufacturing Industry: Chemical Mechanical Polishing,” Master’s Thesis 2007, National Tsing Hua University, Advisor: D. Daniel Sheu (in Chinese).
- Altshuller, Genrich, “Innovation Algorithm: TRIZ, Systematic Innovation and Technical Creativity,” Translated by Lev Shulyak and Steven Rodman; Technical Innovation Center, Inc., http://www.triz.org.
- Altshuller, Genrich, “40 Principles Extended Edition: TRIZ Keys to Innovation,” Technical Innovation Center, Inc., 2005.