Last updated November 1, 1997
Return to TRIZ Page (Home Page)
Archive
Feature Transfer and Trimming
I suppose further, that you have studied tutorials Two Pluses
and Many Pluses.
Invention Machine Lab is a hazardous tool in many ways. First, IMLab donīt give any guarantee,
that the user have selected a right problem. Second, IMLab gives recommendations but doesnīt
evaluate them. Third, IMLab gives different principles and effects, not ready concepts. The
innovative concept is the combination of many principles and physical phenomena. The user
is responsible for the right combination.
A new software changes thinking habits. Driving a car is different from driving a horse. New
ways to think cannot be obtained directly from the software, or learned simply by using
computer programs. Experts who use succesfully Invention Machine are just the people who
can solve problems effectively without computer, too. If someone presents a good idea and
says it was got by Invention Machine, I am absolutely sure that yet something else, additional to
work with IM, was needed.
"Something else"contains many important concepts and skills:
- engineering contradiction
- ideal final result
- monosystem, bisystem, polysystem
- trends of evolution
- psychological inertia
- feature transfer
- function, functional model
- trimming
- ability to decrease temporarily the influence of psychological inertia
- ability to state problems
- ability to evaluate ideas and solutions
Just this "something else" is considered in Two Pluses and Many Pluses.
The most common difficulty the users are speaking of is lack of time. "I have no time to learn the
use of IMLab." There are three solutions to the time problem:
- The first solution is this tutorial which helps to use IMLab most effectively
- The second solution is included in the joke of the man with a bicycle, see Two Pluses
- The third solution: It is not necessary to take in all content of IMLab at once. You can
browse and select small parts: one effect, one principle, one prediction.
IMLab in matrix. Imlab is more easy to understand, if we "insert" the three
modules
in Two Pluses Matrix:
| System | Feature 1 | Feature 2 |
| Object 1 Action Object 2 | + improves | - worsens |
| Alternative system | - worsens | + improves |
| Ideal final result | + improves | + improves |
From the initial system we can select object 1, action and object 2. Prediction gives an array of
predictions of component of the ideal final result. Effects tells how we can deliver the action.
Principles helps to combine incompatible features.
Which module to use? The matrix shows clearly the functions of all three
IMLab
modules. We always have a system which evolves. The system contains always contradictions.
And always the system deliveres some action. The division of labor between modules means
that the best way is to use them all.
Since a detailed analysis by IMLab takes time, we must sometimes choose one of the programs.
In principle the choice is simple. If you know which action your system should deliver, seek
from the Effects module. If you already know how to perform the action but something else
happens that is negative, look at Principles. And if you donīt have a problem but
you would like to know how the system evolves, consult the Prediction.
The users with some experience with IM select without difficulty the module they need. One user
wants work practically only with the Prediction, the another likes the Effects, and the third
thinks all three modules together give the best result.
However, in the very beginning the multiplicity of tools may be a little bit confusing. The Effects
module contains common concepts: action, function, effect. But the action our system is designed
to perform is not necessarily the same one we need to to solve the problem. A matrix in the
Principles contains numerous problem models and standard solutions. It is not always easy
to find a correspondence between the problem and the standard model.
The third module - the Prediction - seems, perhaps, a little complicated, but actually is more
easy to begin from. The Prediction Tree gives with high propability interesting ideas. Thatīs why
I recommend to start from the Prediction.
The example of a bicycle. We would like to know how to improve
the system
"frame-wheel". The model of the system:
- Frame (object 1)
- Fixing (action)
- Wheel (object 2)
- We like to increase the action "fixing"
Weīll look the prediction tree. Some recommendations the software gives:
We looked very small part of recommendations or predictions. The software describes prossible
structural transformations. The userīs job is to interpret general ideas. Maybe elastic substance
(like rubber) helps to make the frame itself flexible enough, so that the suspension is not needed.
Maybe some heat-generating substance allows to heat a bicycle in cold weather. Maybe porous
substances, foames and gels can be used more in bicycles.
We see, that software is not used instead of thinking, but stimulates human brainwork.
The example of a city.The conflict of a car and a pedestrian. In the tutorial
Feature Transfer and Trimming
we stated the problem: How to remove a usual street and transfer its useful functions to
a tunnel or to a highway? Letīs now look at the Prediction. First we build the model of
a problem:
- A tunnel is object 1
- A car is object 2
- An action is supporting
- We like to increase the action "supporting"
We put the description to the matrix. We fulfil the other cells, too:
| System | Ease of use | Mass Size |
| Tunnel supports car | + comfortable | - heavy |
| Usual road | - inconvenient | + light |
| Ideal final result | + comfortable | + light |
Now we can look at different branches of the Tree. Letīs begin from the last part:multiplication.
The reason is, first, that multiplication standards are rather easy to use. Second, multiplication
or the development line mono-bi-poly, is a general law. We can present every innovation as the
combination of known systems and solutions.
How we can improve the action by "multiplying" of tunnels or cars?. Most simple is to use two
or many tunnels or cars instead of one.
The next step: Multiplying and connecting. We get the two-level city we considered in
Two Pluses:
the first floor for cars only, the second one for pedestrians only.
Additionally we got idea of "multiplying and connecting" cars, too. In closed space and by
low speeds is maybe possible to connect cars mechanically or electronically.
The following prediction is multiplying and differing. Different "tunnels", different vehicles. Further:
Multiplying and combining: Some parts of tunnels or cars should be combined. Can many tunnels
or cars have common parts? If some horisontal escalator moves cars, they can, maybe, have
common engine.
More examples of using IM-Prediction:
Exercise 1. Study your own problem. Look at the screen "Trends of evolution".
Do you
find relevant trends, reflecting the past and future of your system?
Exercise 2. Study your own problem. Build the model of a system (objects and
action).
Then transform it by predictions.
We choose corresponding standard features. For example:
- weight of moving object
- shape
We get, for example, following propositions (principles) and examples:
- preliminary action, for example pre-frozen rock
- spheroidality, for example spherical wheel
The next step is ro move from principles and examples to the system considered. Maybe
it is possible to make weak frame which can be easily recovered. Or use spherical or
concave curfaces. A spherical whee is an interesting example. Why not to make
a bicycle with spherical wheels?
The example of a city. Problem statement we get from the matrix:
| System | Ease of use | Mass Size |
| Tunnel supports car | + comfortable | - heavy |
Problem model in Principles:
I WANT TO increase comfort
BY increasing a number or tunnels
BUT THERE IS A PROBLEM mass and size
We choose corresponding standard features. For example:
- ease of operation
- weight of stationary object
We get, for example, following propositions (principles) and examples:
- universality, for example all-purpose gasoline tank
- "the other way around", for example moving the testing ground for tractors
The next step is ro move from principles and examples to the system considered. All-purpose
tunnel? Obviously tunnel which is a part of a building and deliver many functions. Moving
tunnel is maybe less obvious. However, a train carrying cars is rather good analogy of
a moving tunnel.
In Internet you can find much backround information of principles and Altshullerīs matrix.
See articles:
- Z. Royzen. Solving Contradictions in Development of New Generation Products
Using TRIZ, TRIZ Journal February 1997.
- S. Savransky. The Methodology of Inventive Problem Solving, Savransky.
More examples of using IM-Principles:
Exercise 3. Study your own problem. Build a contradiction model.
Then
try different standard models (pairs of standard features). Look at principles list, too.
We seek corresponding standard functions:
- rotate solid
- hold solid
A standard function "rotate solid" solid gives for instance following effects:
- Spirals. An example of a spring motor. Maybe some kind of spring motor
can accumulate the breaking energy?
A standard function "hold solid" gives for instance following effects:
- Ball structures. An example of shaft locking device. Maybe the idea can
be used for improving of brakes.
The example of a city. We want to deliver action "to support". Nearest
standard function is "to hold solid body". IM-Effects suggests for example Archimedesī
principle and Bernoulliīs theorem. There are examples of supporting by water or air.
Maybe for supporting the surface of road will be in future used water and air?
Of using Effects is published in the Net a detailed cased study. Effects for generating small amounts
of electric energy were searched. See article:
- A. Killander. Generating Electricity for Families in Northern Sweden , TRIZ Journal January 1997.
More examples of using IM-Effects:
Exercise 4. Consider the model you built in the exercise 2. Find
corresponding
standard functions. Select useful effects.
We choose corresponding standard features. After two or three attempts we find:
- productivity
- area of stationary object (compare: the area of a street)
We get, for example, proposition (Principles):
- another dimension
Examples:
- mine cars pass in a vertical plan
- vertical boat storage
"Mine cars" give an interesting association: why not to use highly developed mining technology in
building of cities?
We will now look at the third modul, Effects. We want to deliver action "to carry cargo".
Nearest standard function is "move solid body". IM suggests 29 effects, for example Bernoulliīs
Maybe two-level city allows to use some new physical principles in transportation?
Valeri Sushkov has compared TRIZ with second-generation soundcards. We have a large
collection of data (examples), and simple algorithm. We can compare the data with a map.
The simple algorithm is a "compass". See:
- A Sushkov. Technosocial Competition, TRIZ Journal February 1997
Separation in space:
Many principles and effects can solve more than one type of conflict. Segmentation is so
important principle since it allows separate features and properties as well in space,
as in time and structure.
So, before working with IMLab is useful to define where the conflict lies. When we know that
we should separate properties of a street in space, we need look at only information related
to space: segmentation, multiplication, "nested doll" and so on.
Or, for example, we need the package material certain time. After that time the package is
useless and harmful waste. Conflict in time. IMLab contains several ideas how to control
properties in time.
At last, we can have a situation when conflicting properties are required at the same time
in the same place. For example: low weight and high strengh. Or good compressive
strengh and high tensile strengh. Solutions of this like problems have one common
feature: the system as whole and the parts of the system have different properties.
For example foam, porous materials and composite materials.
Last I would like recommend to look at manuals.
Return to TRIZ Page (Home Page)
Archive
Feature Transfer and Trimming