Generator - Forecasting Maps of Engineering System Evolution

To create a promising product, it is necessary to forecast, with the maximal accuracy, the future situation for similar engineering systems.

How is a forecast of engineering system evolution made? It is normally made by the method of extrapolation. The past evolution of an engineering system is examined and its place on the S-shaped evolution curve is determined (Fig. 1). Then a conclusion is made about a probable variant of its evolution taking into account the evolution trends of the engineering system and the parameters to be improved. This information is usually not enough and an adequate forecast is not easy to make by this method.

One may draw an analogy with the exploration of minerals. If we want to find some mineral and we have already found it in mountains, this does not mean at all that we will find it in any mountain. For an effective exploration, it is necessary to compile a detailed map of an explored region, to mark all peculiarities of its topography and earth crust structure, to determine and mark magnetic fields, climate peculiarities, to ask around where lightning often strikes during thunder-storms, and to check what kinds of plants grow there. Then a conclusion can be made: the occurrence of this mineral is highly probable in one area and unlikely in another area. The main document for such a forecast is the map of the explored region that contains the fullest possible information.

Forecasting the evolution of an engineering system has much in common with this situation. A great help will be full information in the form of a table or a map describing all basic embodiments of a given engineering system, both realized and non-realized. If we know the current state of an engineering system and understand which operational parameters of the engineering system have already reached their limit and which of them can be and must be improved (productivity, speed, operation precision, etc.), we will be able to select, with a great degree of credibility, the versions of the engineering system which are likely to appear in future, that is, to forecast its evolution.

Any engineering system has its main useful function. At the same time, any engineering system can be presented as a set of simplest, elementary functions performed in correlation and in a certain sequence. If we could calculate all possible methods of performing those functions, we would create a comprehensive model of the technical world and thereby obtain material for compiling forecasting maps. The simplest, elementary function is a derivative of some finished useful action that cannot be segmented any more into qualitatively different functions. The examples of those functions are “to cut an object”, “to move an object”, “to irradiate an object”, “to cool an object” and the like.
To describe all possible methods of performing an elementary function, it is necessary to present it in the form of an elementary action in the operational zone. To perform the action, it is necessary to have a tool (an object that performs the action) and a processed object (that undergoes this action). Besides, the environment must be taken into account.

For instance, Fig.2 presents the performance of the function “to clean surface of object”. It comprises 3 elements:
1. A tool that performs the main working motion (scraper).
2. A processed object (the surface to be cleaned).
3. Environment.

Then we should describe all possible transformations of each element of the operational zone. The simplest well-known “transformer” is the morphological box. Its use in TRIZ is very limited because it produces a great number of variants without pointing out relevant solutions. Another disadvantage of the morphological box consists in that while building it, the features of an element or system under research are arranged along its axis randomly. This results in subjective and incomplete selection of transformation variants.
B. Goldovsky [1] proposed follow idea.
In our case, the vast number of produced variants is rather an advantage than a difficulty, because we must cover the entire field of transformations of an element. To avoid “blind-spots” in the produced map, it is necessary to use some objective checkpoint. For the technical world such a checkpoint is the trends of engineering systems evolution. We can arrange not a random set of features along the axes of the morphological structure, but transformation variants of a given element fulfilled in accordance with evolution trends. Since there are more than three lines, we can talk about a multidimensional morphological box or about a transformation tree of an element.

We can demonstrate the method by building a fragment of the transformation tree for the function “Clean surface of object”:
The starting point is the variant, which is the simplest one in the context of the trends of ES evolution – a solid rigid object with flat surfaces and without internal structure – a rigid scraper (Fig.3).
The scraper evolution line is laid off along the vertical axis. It is built in accordance with the “Object segmentation” trend. This is the trunk of the future tree. The solid scraper is segmented into several narrower blades along the vertical axis and turns into a brush, a jet of sand, sticky substance, liquid, foam, gas or plasma. Then it is segmented to the field level and further on to the vacuum level.
The horizontal lateral branches-lines are laid off each version of the "segmented' scraper. Those lines describe the evolution of a given version of the scraper. The number and composition of those lines are determined by resources. For instance, the important resource for the scraper is its cutting edge and front surface. So it is expedient to add the lines “Geometrical evolution of linear constructions” and “Surface segmentation” to the scraper evolution lines.

introduction of a new substance into a scraper;
mono-bi-poly – two and then several scrapers that follow each other;
dynamization by introducing flexible ties – up to a flexible scraper and multi-hinge chain or band “cleaner”;
dynamization by introducing an additional action – vibration, ultrasound, etc.;
coordinated action – resonance phenomena;
activating by increasing the speed and force of the scraper pressing, changing its temperature and other necessary parameters;
geometrical evolution of the scraper blade shape, for instance, making the blade wavy;
geometrical evolution of the front surface of the scraper – making it concave in one and then in two directions for collecting impurities;
segmentation of the front surface of the scraper, forming guiding protrusions and roughness;
segmentation of the scraper volume, for instance, introduction of a void with the supply of liquid or air into the near-the-blade area.

The scraper segmented to the “liquid” state:

introducing new and modified substances to a liquid jet – abrasive particles, pieces of ice, gas or vapor bubbles, chemically active substances, etc.;
mono-bi-poly – two liquid jets, several jets, multi-jet structure;
dynamization – complicating the motion trajectory of a liquid jet, generating pulsation and ultrasound, rotating liquid in a jet;
coordinated action – resonance phenomena;
activation – increasing the speed, changing the temperature of liquid, focusing and defocusing the jet;
geometrical evolution of the liquid jet surface – making it flat, concave or convex, or giving it the form of a hollow cylinder;
evolution of the contact spot shape as a transfer from a point to lines of different curvature;
segmentation of the jet volume, transfer to jet atomization and mist generation.

Thus, we obtain a structured list of possible transformations of a selected element. To obtain all possible variants of performing a given function, it is necessary to select in turn an appropriate version of a tool and a version of a processed object and environment in which an action occurs. Transformations for one and the same element can be combined. The selected transformed elements are combined to demonstrate one of the possible variants of function performance (Fig. 4).

In this case a new, unexpected effect often occurs. For instance, if we choose an air flow as a tool for cleaning the surface and if this flow is parallel to the surface, the cleaning effect will be insignificant. But if we bend the surface as a part of a cylinder and expose it to that flow, then aerodynamic force will occur on its convex side. It will be similar to the force occurring on the airplane wing. The separation of contaminants will improve abruptly (Fig. 5).

The number of such scenes is very large and is expressed by a number to the 9th power. At first sight, it is not easy to orientate in this array of information. However, the number of variants reduces abruptly while making a real forecast. In this case we are interested in the element transformations aimed at the improvement of a required operational parameter. Hence, we lay an emphasis only on some trends. Thanks to this approach the number of considered variants reduces abruptly. In addition, we can limit the number of variants by applying an additional filter in the form of a substance-field resources available in a given invention situation. In this case the forecast takes into account first of all the resources available for use in a given Engineering System. With such an approach, we obtain a considerably small amount of information grouped by evolution trends, which helps us make a more accurate forecast.

In addition to being used directly in making evolution forecast, such a forecast map is extremely convenient in bypassing patents and creating “patent umbrellas”.

References:

1. B.Goldovsky. TRIZ in the rapidly shifting world. About different approaches to subject TRIZ. TRIZ Journal. C.3. N 3.1.92 (In Russian).