The duality of Altshuller's view on ideality

Y. B. Karasik
Thoughts Guiding Systems Corp.
Ottawa, Canada.

Altshuller gave two definitions of ideal machine:

  1. Ideal machine is a machine which has almost the same weight, area, and volume as the weight, area, and volume of the object that it transports, processes, etc., i.e. works with. Also the ideal machine and all its parts never idle but fully engaged and work to the fullest [1].
  2. Ideal machine is when there is no machine but the result is the same as with a machine [2].

As an example of the ideal machine he mentioned a truck with zero weight and just a cargo compartment, no cabin, etc. But such an example does not match the definition. Neither weight, nor area, nor volume of such an ideal truck coincides with weight, or area, or volume of cargo. The weight of the truck is zero whereas the weight of cargo is not. The area and volume of the cargo compartment could be made equal to the area and volume of the cargo only if it is a cube van. But if it is a flat bed then area and volume of the compatment is less than the area and volume of the cargo.

Thus the first definition is inconsistent. The consistent definition would be this:

But it is almost the same as the second definition. Zero weight, area, and volume exactly means that there is no machine but its function is fulfilled. Thus, we can stick with the second definition as the only consistent one.

Altshuller also claimed that all machines evolve towards the ideal machine, but did not specify whether they EVENTUALLY evolve into the ideal machine or any step in a machine's evolution brings it closer to the ideal one. Meanwhile there is a big difference between the two concepts of evolution.

If any machine just eventually evolves into the ideal machine then the only important transformation is its final leap to the ideal state. All previous transformations are unimportant from this prospective as they do not necessarily bring the machine closer to the ideal one. But since any transformation of a machine is its improvement in a sense, it means that improvements do not necessarily make machines more ideal. In this case trying and defining the degree of deviation of a real machine from the ideal one does not make any sense. It may show no direction of the required improvement of the machine at the moment.

By the way, the purpose of ARIZ is also not to somewhat increase the ideality of a system but to convert it into the ideal one in one quantum leap. To this end the Ideal Final Result (IFR) is formulated and then is tried to be achieved. Failure to achive IFR is considered to be a failure to solve a problem by ARIZ. No wonder that the measures of the degree of ideality of a machine found no place in ARIZ.

On the other hand Altshuller once wrote that the fact that the maximum size of trucks grew over time, does not contradict the tendency to ideality. The ratio of the load to the truck's size has been lowered. Thus in this instance he assumed that machines not only eventually become ideal but that every step in their evolution makes them more ideal. Still in other instances he used to say the opposite. Here is an excerpt from the recording of his argument with some of his disciples that were pushing him to define the degree of ideality [3]:

It is easily seen that in this particular instance Altshuller meant that machines eventually become ideal and that their intermediate improvements may make them less ideal. Thus, his views were quite dual, although the evidence suggests that he adopted the latter point of view more often than the former.

And if improvements do not necessarily make machine more ideal (until it finally leaps to its ideal state), then there is no need in formulas of ideality. Indeed, what is the purpose of such formulas ? It is to measure how far a real machine is from the ideal one and direct it towards the ideal one. But as Altshuller pointed out in the above discussion, it may turn out to be a wrong direction of improvement. Thus, formulas of ideality are useless at best and misleading at worst.

Instead of inventing the measures of the degree of ideality it is better to invent other useful measures of the quality of machines. For one that any improvement increases some quality of a machine although not necessarily the degree of ideality. In this respect degree of ideality is no different from other qualities which may or may not be improved. It is just one of the possible qualities although so far no one was able to propose its consistent measure. Thus we need to identify all qualities that might be improved. One such new quality measure for transportation systems was proposed in [4] and more are on the way.

R E F E R E N C E S:

  1. Altshuller, Genrikh. The algorithm of inventing. 2nd ed. Moscow: The Moscow Worker, 1973. 81. Print.
  2. Altshuller, Genrikh. The algorithm of inventing. 2nd ed. Moscow: The Moscow Worker, 1973. 83. Print.
  3. Published on the internet at various web sites (see e.g.
  4. Karasik, Yevgeny. Do systems really evolve along the S-curve ?. Anti TRIZ-Journal, Vol. 4 No. 5, October 2005.