LAW OF REQUISITE VARIETY


The law of requisite variety originates in the field of cybernetics, control and system theory. Cybernetics was defined as “the science of control and communication, in the animal and the machines”. The law has special significance in the development of engineering control and communication systems but it’s a general principle that applies to any system, whether economic, social, mechanical or biological.
Ashby’s Law was derived from mathematical analysis, but in plain language it is very simple and seems little more than common sense. According to him, control can only be obtained if the internal regulatory mechanism of a system is as diverse as the environment with which it interacts. Note that the variety is requisite, if the control system is too complex the system will not operate efficiently; but if it lacks sufficient internal differentiation, it might not be able to cope with variable demand, or the system might fail entirely. One interpretation of this rule is that adaptive, flexible systems which generate a large range of diverse options are better able to manage change than systems which are tightly optimized around a rigid set of initial conditions.
The variety must be at least as great as the variety of disturbances. Ashby has called this principle the law of requisite variety: in active regulation ‘only variety can destroy variety’. ‘Variety’ in mathematical term is defined as the total number of possible states of a system.
The law of requisite variety implies that the selection of measurements, as perceived by all relevant individuals, is a fundamental aspect of any management system. The law states that in order to regulate a system, the regulator must be able to generate at least as much variety as that exhibited by the system under control.
The law of requisite variety states that in order to control each possible states of the system elements, there must be a corresponding control state. Suppose if there are 50 items to control, there should be at least 50 control states. To put it in another way complete control of a system can be achieved only when the control system has as much variety in response as the number of ways the system can go wrong. This means that the controller of a system must be able to determine the variations of the control variables and required to sent system changes instructions for each change. If there is large number of items to be controlled, it is humanly impossible to generate control responses for each possible variation taken place in each item.
The management can solve the above problem by formulating standard control responses that can be commonly applied to all items. In addition to this, the subordinates are supplied with the decision rules for generating required control responses in special case.
According to this law, in order to control a system, the system controller must be provided with the following elements:
  1. Enough control responses to make the controller capable of handling any situation.
  2. Decision rules for generating all possible control responses.
  3. The authority to become self organizing system in order to generate control responses.
The most suitable mechanism for achieving control is to set up a man machine system. Computer system can produce the required control responses for all expected situations. As far as unexpected situation is concerned, the human decision maker can apply his own brain and commonsense to achieve control.

Post a Comment

0 Comments