Last Modified:                                                                                                



Homepage / Publications & Opinion / Silicon.com

Peter Cochrane's Uncommon Sense: Forget 'think different' - think 'non-linear'...
Let's start applying 'deep thought' to some everyday problems...

For years I have been trying to convey to educated and lay audiences the key differences between a world that is linear and one that is not. This might in principle seem a simple task but it turns out to be incredibly difficult. The fact is that most of our education and worldly experience is founded on linear assumptions and from an early age we become conditioned to a majority experience that leads us to believe it is the dominant condition. But the converse is true!

Of late I have been working with organisations that have always optimised their operations along traditional lines with singular cost and performance parameters receiving all the attention. Without exception their single biggest mistake has been to optimise the individual components of a serial process and then glue them together end to end as a complete system in the belief that the whole will then be optimal. But this is seldom the case. Why?

Most of the time we work with materials and systems in bounded and well-controlled environments. For example, if we stretch an elastic band it will expand in direct proportion to the force being applied. But should we go beyond 'the elastic limit' it will suffer damage and ultimately break. The same is true of wood, metals and plastics. Most of our design work is concerned with materials subjected to given tension and compression that will see them return to their original state after being flexed. We purposely concern ourselves with supporting structures that will not be damaged by vibration, such as the framework of a building, bridge or aircraft wing.

If we were designing an automobile and wanted to make the drive chain as efficient as possible we could work independently on the gear box to reduce the friction and optimise the gear ratios for acceleration and top speed and similarly the back axel and differential. We would ensure the design of the engine gave maximum power for a given fuel consumption. When all were linked together to constitute the whole drive chain, it would indeed be optimum, providing a few issues of energy and torque matching have been taken care of.

But this is an inherently linear system. The energy output varies in direct proportion to the energy input given the losses due to frictional heating and noise. In a limiting case we could rev up the engine to the point where the materials overheat, extending beyond their normal limits, and then the system would become non-linear and irrevocably damaged. But the fundamental design assumes all of the materials and the operation will be sustained within the limits of linear operation. This example is as good as any in terms of defining our linear thinking and approach to optimisation.

So where does it all go wrong? Almost all resource and logistics problems are inherently non-linear. Probably one of the most graphic examples today would be the hospital bed manager in the UK or the bed marshal in the US, employed to achieve bed fills in excess of 99 per cent, which they do with great regularity. Unfortunately, this single point efficiency leads to an overall system inefficiency and is a significant damage generation mechanism.

The first effect is due to the lack of time for bacteria to be cleared between bed occupancy - which leads to cross-infection, longer treatment times and patient return visits.

Second, surgeons who would normally have all their patients grouped on a single ward or known location now have to locate them in wards and rooms randomly distributed across an entire hospital. In the worst cases patients may be moved from ward to ward on an almost daily basis. So overall surgeon and general medical staff productivity goes down.

Third, transport, administration and overall operating costs escalate as medical records, patients and medics become disassociated. And of course, people become upset and dissatisfied as the inefficiency is all too obvious.

Fourth - and a ticking time bomb - the hospital system is rendered unresponsive to any crisis of any scale. So an air crash, major road accident or terrorist event leads to chaos spanning many hospitals and the road network, as patients and victims are transported to the nearest facility with available beds.

The really damaging outcome is that people die, people don't get cured and people become dependant upon long-term care. Patient dwell times are extended, the overall system is made more inefficient, costs escalate, satisfaction goes down, politicians get involved and a downward spiral begins.

The answer? Joined up thinking. A full and all-encompassing analysis where the optimisation is end to end, from in to out, start to finish - and not a single point function.

I think I should make it plain that I am not singling out health care in particular. The above observations are equally valid for transport systems - public and private, road and rail, containers and parcels, atoms and bits. And that's not to mention government, management, ecology, energy and living.

Everywhere you look there are huge savings to be made by the application of deep thought, true understanding and effective models.

This column was typed on my G4 laptop flying Chicago to London on BA298 and despatched from my home on a super-fast high-speed connection a couple of weeks later.