The Viable System Paradigm

The Viable System Model was developed by Stafford Beer from his neuroanalytic T, U and V machines produced to construct a "Cybernetic Factory".

Ashby's pioneering work on variety and homeostasis and Pask's work in electrochemical concept growth are important citations, for example.The Beer VSM paradigm is relevant as ever to developing Nano Technology Assemblers and Replicators, for example, because of its deep insights into Organisation in the face of almost inconceivable complexity.

Thus Beer founded Management Cybernetics.

The Five parts correspond to Decision or Identity (System Five) Development (System Four), Operation (System Three), Regulation (System Two) and System One where the whole structure repeats. Systems 3-2-1 form an Autonomic Loop with Four and Five corresponding to cognitive and higher mental functions.

Stafford's core idea was to take human neurophysiology and make an abstraction that mapped into the business or industrial process. The relevant statistic drives his models with actual performance statistics compared to capabilities and potentials.

• If Actuality deviates from Capability too much an exception signal is sent to the responsible manager.
• An index of Potential established from current Development Plans and representing what can be achieved if the plan is implemented.
• Capability is what can be done on a good day when everything is working as it should- a standard.
• Alerting Algedonic Feedback- analogous to pain in an organism- tells senior 5-4-3 management when life threatening events occur. In the Public Service, for example, when performance is non-compliant with a service delivery agreement a manager is informed. This principle is recursive: it repeats throughout the Organisation.

A Tale of Feedback and Management Cybernetics

There once was a rolling mill.

Very skilled men controlled it with a variety of techniques virtually inherited father to son. The thickness of the steel however was rather variable and if the wrong man was absent quality control went out of the window.

The Board decided to buy a computer to automate control of the process.

Experts were consulted and mathematical models constructed. Many parameters were found to control the thickness of the steel coming off the rollers. Ambient temperature, steel temperature, speed, pressure on the rollers, composition of the steel. Much sensing at multiple points around the mill and many rollers was contemplated. Much summing and weighting of outputs to control speed, room temperature, roller speed, pressure and gas burners heating steel about to be rolled.

The computer was late. Further heads were scratched and the idea came of directly monitoring the thickness of the steel then using that output to feedback to ANY of the controlling variables eg gas flow to pre-heating burners. That is the power and elegance of correctly applied negative feedback. They solved their problem with £5.00 worth of electronics.

When the computer arrived they used it for the payroll.

The Great Failing in Management today in large Enterprises is to measure everything and pass it to the senior management. The figures are then contemplated and strategic instructions are issued. How would the rolling mill function if this was how operations were managed? The people on the ground are kept in the dark instead of seeing immediately how their tactics are working. Proper timely feedback is all that is needed. We commend to you the work of the Real Time Study Group.

Stafford Beer