The swarm as a method of work organisation

Excerpted from Bob Cannell:

“It is possible to organise communications in a better way. One that utilises the collected intelligence of the agents (people) in the network rather than suppressing intelligence into the obedience necessary to operate a business process production line (and then moaning about your workers for not obeying the rules).

The most serious problem with Fordist production lines is resistance to change, the investment of effort required to change is high, and it therefore lags behind customer and market changes. Change has to be reactive because it has to be planned and agreed in advance (i.e. not worth the hassle).

Fordism is fragile because it is dependent on experts and specialists who alone can control it and it is vulnerable to disruption, being linear, a blockage stops the line. (modern multi-lateral production lines like Dell’s avoid the ‘spanner in the works’ blockage but are still vulnerable to shortages of key components produced by single factories). You don’t have to have a physical production line to be Fordist, it can be virtual or spread out between departments or between sites. It can be manufacturing or service.

Fordist production dehumanises people working in it, alienating them from their behaviour, generating stress and negative reactions, and divides people into controllers and controlled. Worker coops struggle to manage these tensions.

Organisation based on swarm intelligence is different. It looks to the ‘intelligent’ behaviour of social insects, where non-intelligent single insects working socially, can create extraordinarily complex organisations – two way highways, brutally efficient foraging behaviour, organised defence strategies, bridges over obstacles, complicated division of labour and large and complicated physical structures like termite nests.

These organisations look like factories. The behaviour looks like efficient productive behaviour. It is not reliant on single pathways or single specialists or single resources.

To our Fordist eyes there seems to be redundancy. Ants exploring seem to wander aimlessly though they are actually following a search pattern. Once a food source or threat or job needing doing (clearing out the rubbish for example) is detected, the information is efficiently transmitted and colleagues easily exploit the new found resource, mount a defence or get it done. Opportunities are not wasted as they are in Fordist organisation where people (not ants) cannot try something new because they are too busy ‘on the line’ or ‘on the phones’.

This is the key weakness in Fordism, that people’s intelligence is not used, yet business managers cry out for better change capability while the source of improvement is there in front of them but frustrated by the organisation they impose!

Swarm intelligence based production systems are not as dehumanising as Fordist production lines. Strange as it may appear when we are looking at ant behaviour, swarm permits people to use their intelligence to try a new way of doing things, or try new possibilities. If they are supported by appropriate information systems (the human equivalent of the ants’ pheromones) successful ‘tries’ are efficiently copied and the effort gets a good payback, and unsuccessful ‘tries’ get stopped PDQ.

But the key is that the ‘changes’ are small and build up to a bigger change if other people copy them. So production systems cannot be production lines where even small changes cause disruption. They have to be something like production groups eg as Volvo (used to) use, a team of workers builds the car in a bay with parts brought to them as needed, rather than standing alongside a line with parts brought to that stage on the line.

In the group, workers can show team colleagues better ways to do the work, peer review encourages quality and collectively demand better support from others eg parts distribution. On a Ford line, workers face the machine as individuals unless they all go out on strike together. There is little possibility of self-managed improvement (thought there may be suggestion boxes managed by management.) They are dependent on management to make improvements , to respond to feedback, to be persuaded to give them the time and space to try something new etc.

In a swarm organisation, this creative behaviour just happens. HR recruits, selects and trains people to work in this self-managed way but we then force ourselves to work ‘on the line’ instead.

Swarm isn’t chaos, there are principles and boundaries. Ants normally behave predictably, but if something different happens, maybe they fall off a leaf, they can behave differently. Humans being intelligent are much better at taking initiative than ants but normally most people prefer a reasonable routine with an occasional challenge, just right for swarm organisation.

It would be easy to reorganise worker owned businesses as we want to, to fit how we clearly would prefer to work, in charge of the work rather than the work being in charge of us. The reason we don’t do it, is because our information and business process management systems cannot currently cope with anything more complex and we assume the generic production line is the best and only way to organise.

So if we want our businesses to be more responsive to customers’ changing requirements, automatically correct errors, based on team working rather than ‘cloud’ working (much work is a bit like being a wandering lonely cloud), which enables rather than inhibits continuous improvement, should we be looking at organising ourselves something like swarm networking or like a production line with pre-set rules? And shouldnt any new IT system we are designing, be designed to support swarm networking rather than to control a production line (because then we would still be stuck where we are).

Too often more modern systems analysis opportunities are not considered before the trusted old ways are chosen again. Value Chain Analysis and other similar Business Process Re-engineering methods were invented long before the internet so it’s not surprising they are designed to improve linear or simple networked processes.

It’s not surprising they focus on designing a production process into which people must fit (requiring management to make them behave according to the rules of the process, an unwinnable task in a worker owned business).

And its not surprising they create controllers and other people who are controlled (even if they democratically agree to be controlled) because that is what they are designed to do. A 2006 European study found the primary cause of degeneration of worker coops was capture by experts who come to dominate and control information. Creating controllers is not safe in worker owned or cooperative business.

Should we not be looking at 21st century forms of organisation and creating computer systems to support them rather than assuming the only option is the old way?”

1 Comment The swarm as a method of work organisation

  1. AvatarPoor Richard

    “A 2006 European study found the primary cause of degeneration of worker coops was capture by experts who come to dominate and control information. Creating controllers is not safe in worker owned or cooperative business.”

    This is an interesting observation and I think there may be an important issue to explore.

    Humans share many genes with other social animals. One thing we can observe in many social species is the way that “status” genes can be turned on by social circumstances. In many species when an “alpha” individual is lost by the pack or herd, a formerly subordinate individual will fill that role. Not only does the behavior of such an individual change, but in many cases there are physiological and morphological changes that can accompany such status changes even in fully developed adult individuals. This may be mediated by epi-genetic mechanisms.

    It may be that humans (perhaps some more than others) are similar. Put some people into a group of cooperating peers where there is no alpha individual and this may actually trigger something in them to assume an alpha role.

    In humans it is especially difficult to distinguish between psychological and genetic triggers for behavior, and my point is not to make a case for genetic determinism. I am only suggesting that the variety of unconscious and involuntary forces that might affect human behavior can run very, very deep.

    If leaders, controllers, etc. are dangerous for cooperative peer groups, it may take a lot more than wishful thinking to find ways to suppress the tendency of humans to express such phenotypes.

    It occurs to me that we might try to incorporate stimuli in the workplace that would somehow inhibit any tendency for alpha traits to emerge and drive individuals to fill status roles that are vacant by intent–if there were some kind of artificial “decoy” alpha in the room, for example.

Leave A Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.