The TRIZ Journal

	Bookmark This Page Email This Page Format for Printing

Assessing The Accuracy Of The Contradiction Matrix For Recent Mechanical Inventions

[ Previous ] [ Home Page ] [ Up 1 Level ] [ Next ]

Assessing The Accuracy Of The Contradiction Matrix For Recent Mechanical Inventions

Darrell Mann
Industrial Fellow, University of Bath
Bath, BA2 7AY, UK
Phone: +44 (1225) 826465
Fax: +44 (1225) 826928
E-mail: D.L.Mann@bath.ac.uk

While some parts of the TRIZ community might debate the value of the Contradiction Matrix (Reference 1, 2) as a problem-solving tool, it is undoubtedly an attractive concept to both users and newcomers. A large part of the dis-satisfaction with the Matrix seems actually to stem from the fact that it is out of date rather than conceptually incorrect, and that to update it would require a lot of effort that might be better spent elsewhere.

A previous article (Reference 3) included the following figure as a way of highlighting the extent of the problems associated with the current classic Matrix. Or rather, due to its lack of quantified data, suggested the extent of the problems.

This article reports a brief study conducted by 32 final year engineering and design undergraduates at the University of Bath to calibrate the Matrix against a cluster of recent mechanically-oriented inventions.

The study was conducted by allowing students to select a random set of patents from the US and European patent databases, the only criteria being that the patents should describe an essentially mechanical-based system and that the patent should have been granted in the last ten and preferably five years. No other selection criteria were formally applied, although there was an overall suggested direction towards picking inventions that were ‘interesting’ in some form or other - this usually meant a solution featuring some kind of ‘wow’ moment. Each student was tasked with identifying four patents for analysis - and hence the total dataset for analysis comprised over 130 different inventions.

For each patent, the student was required to identify what aspects of a design the inventor was seeking to improve, what parameters these aspects conflicted with, and how the inventor overcame the conflict. In broad terms, these elements can be extracted from the background, summary and claims sections of the invention disclosure text. For each patent, the student was required to cut and paste the relevant section of text, show how it mapped onto the terms of the Matrix, and show if/how the inventive step related to one or more of the Inventive Principles. There was no instruction to distill either conflict parameters or Inventive Principles down to a ‘best one’; if the inventor solved more than one conflict, or if the conflict was blurred across several parameters, or if more than one Principle was used, the student was asked to record all of the information.

Below is an example of the type of analysis conducted for each patent:

From the patent text…..

BACKGROUND OF THE INVENTION
The invention relates to a device for applying labels, as defined in the preamble of the independent claim 1.

Applying labels to flat mail items quickly and reliably poses a problem in the processing of flat mail items,

particularly letters, post cards, etc., in mail-processing facilities. An example of this is automatic mail forwarding.

In this instance, mail items are separated out for forwarding addressing, and re-addressed with corresponding,

predetermined data stored in a database. A label that covers the old address and, if applicable, a barcode

printed on the surface of the mail item, is affixed to these items. The label is then provided with a new barcode

and the corresponding, new address. The label is applied in devices that are integrated into automatic letter-

distribution systems. The mail volume for such distribution systems varies in format, weight and thickness. In

these systems, the items are conveyed at a speed of, for example, 3.6 m/sec, which places stringent

requirements on the speed at which the labels must be applied, as well as on the exact positioning of the

labels. Furthermore, the handling and especially the transport of the labels to the item surface represent a

general problem when the label has a self-adhesive surface.

A device of the above-outlined type is disclosed in U.S. Pat. No. 5,200,007.

It is the object of the present invention to provide a device for applying labels to flat mail items, wherein the

supply of the labels to the region of application and the trimming of the labels is performed inexpensively with

high advancing speed and accurate positioning.

According to the invention, this object is accomplished by the features of claim 1. Advantageous embodiments

of the invention ensue from the dependent claims and the description.

A notable advantage of the invention is a high flexibility in the positioning of the labels on the mail items, and

a simple process of supplying the labels to the application region. The conveying speed of the mail items

need not be reduced for label application. For the duration of the process of supplying the labels to the

application region, and during the application itself, the labels are mechanically held, so no uncontrolled

changes in position occur. The labels can be supplied from a simple label strip or a label carrier strip.

etc

….the student is required to pick out the conflicting parameters….

….see what the classic Matrix recommends as its ‘top 4 most likely’….

….and then pick out the Inventive Principles actually used by the inventor….

….and so, in this case, we are able to record a partial match between Matrix and what the inventor used. Thus, in terms of quantifying the ‘success’ of the classical Matrix, this invention scores one out of two - i.e. two Principles were used; the Matrix suggested one of them.

The total success rate of the Matrix was then calculated by summing all the Principles contained in the Matrix used by each inventor, and dividing that total by the sum of every Principle used in each of the patents under evaluation.

Throughout the exercise - carried out over a period of two weeks - students were also asked to pay attention to either conflict parameters or inventive strategies that did not fit into the existing Matrix/Inventive Principle framework.

The following table summarises the analyses for each of the patents successfully analysed. In order to ensure consistency of analysis across each patent, this author has monitored each one individually. The table provides patent number, title, improving parameter(s), worsening parameter(s), Inventive Principles recommended by the classic Matrix, and Inventive Principles used by the inventor. Anyone wishing to see the specific analysis for any of the patents in question may request a copy from the author. Alternatively, you may like to conduct a few analyses for yourself to see if you agree with the diagnoses presented here.

Of the patents analysed, the net effectiveness of the Matrix was calculated as 48%. In other words, in 48% of inventive problem situations, the Matrix in its current form would have enabled the inventor to reach the eventually patented solution.

The classical TRIZ Contradiction Matrix was assembled from primarily mechanically biased patents from twenty or more years ago. While the sample here is of a significantly smaller size than the original analysis, it is nevertheless sufficient to allow some statistical comparison to be made. The 48% success rate, although lower than some estimates (higher than others, however) should give us at least some confidence that the Matrix is still relevant for mechanical problems. The most significant shifts in thinking appear to come through increased use of Principle 23, Feedback, and to a lesser extent 25, Self-Service - and these now appear to be somewhat under-represented in the classic Matrix. Conversely, while the Matrix suggests Principle 22, Blessing In Disguise relatively frequently, the patents analysed suggest it has rarely been utilized. In terms of the 39 parameters contained in the Matrix, it was evident throughout the monitoring of the patents that parameter 31’Object Generated Harmful Effects’ was used in a relatively high number of cases. The parameter has to be used to wrap up a wide variety of issues (safety, noise, environmental harm, etc), which would ideally be handled by introducing more parameters into the Matrix.

The classical Matrix, of course, should not be interpreted as saying that the Principles used or usable to solve a given contradiction type are limited to the ones published - merely that these are the ‘most likely’. If the Matrix recommends any three or four Principles for a given contradiction, it does not preclude the achievement of a strong solution from one of the other 36 or 37 Principles. The results of this study are suggesting that, for mechanically oriented problems, the Matrix recommendations will be ‘right’ just under half of the time. We should continue, therefore, to use the Matrix as a ‘useful start point’ rather than as a definitive end.

If TRIZ is about finding win-win solutions and maximizing the utilization of resources, the idea of using students to help update the TRIZ methodology appears to offer a very good match. The activity conducted by this batch of students served to familiarize them with searching patents, observing the way patents are written, understanding contradictions, familiarizing themselves with the Inventive Principles, in addition to providing the TRIZ community with some hopefully useful data.

It would be good to extend this type of analysis to patents from non-mechanical areas in the future. The author is happy to pass on teaching and assignment instruction data for anyone wishing to attempt a similar exercise with students elsewhere.

Meanwhile, CREAX nv in Belgium is in the process of systematically analyzing and codifying patents for contradictions, trends of evolution, inventive standards and knowledge/effects content, starting from the present day and working back in numerical order, patent-by-patent back to the time when systematic analysis previously stopped. The fruits of this updating activity will gradually be published via www.creax.com. Anyone wishing to find out more about this work, or wishing to participate in some way should get in touch with the author.

The author would like to thank all of the students for participating in the exercise; without their efforts, this article literally would not have been possible.

1. Domb, E., ‘Contradiction Matrix’, TRIZ Journal, September 1998.
2. Salamatov, Y., ‘TRIZ: The Right Solution At The Right Time’, Insytec BV, Netherlands, 1999.
3. Mann, D.L., ‘The Space Between Generic and Specific Solution’, TRIZ Journal, August 2001.

ã2002, D.L.Mann, all rights reserved.