This form of team working, using a computer, raises some interesting and important questions in computer science. How, for example, can such tasks be arranged in the optimal schedule?
To address such questions, researchers are borrowing techniques from biology; in particular from work carried out by those who are studying the group behaviour of insects such as domestic bees.
A typical application for this research is in an area known as concurrent engineering. This is a branch of conventional engineering that is attempting to shorten the lead time required to develop a manufactured product. A company that can release a product faster than its rivals is able to achieve a competitive edge that often lasts for the lifetime of that product. In the car industry, for example, a vehicle manufacturer who releases a new model six months earlier than its competitors can increase its profits by tens of millions of pounds on that model.
In conventional engineering, producing a new product involves a series of steps that are carried out one after another: product specification, product design, manufacturing design and assembly. To speed up the engineering process, concurrent engineering replaces this sequential process with a series of parallel activities that are, as far as possible, carried out at the same time. Unfortunately, developing a product such as a car or a video recorder is a complex process, and a computer is needed to carry out managerial activities such as assigning tasks to engineers, updating engineers when a specification changes and monitoring progress.
Concurrent engineering and other promising areas such as computer supported conferencing, which uses computers to co-ordinate the actions of large teams, are potentially powerful technologies. Unfortunately, major research questions are hindering their spread. For example, there are questions about how the limited amount of human resource on a project can be optimally scheduled in order to minimise delivery time.
There are also questions about the best way to organise a large team that works concurrently: the normal way of organising a project - as a hierarchy of seniority - is not the best way to use staff when there is a high degree of parallelism to the tasks which they have to carry out.
A number of disciplines are being used to solve these problems, including economics, organisation theory, sociology, computer science and cybernetics. However, the most interesting involves the application of field work carried out on insect and animal behaviour.
What attracts researchers to biology is that very simple rules, almost trivial to implement by a computer, enable a collection of animals or insects to carry out very complex tasks. This elevates their collective skills to the point where their behaviour is like that of a super organism. For example, researchers at the Massachusetts Institute of Technology and Yale University have been using observations carried out on lion prides to study resource allocation in computer-supported work. They are attempting to translate into task-allocation terms the observation that prides of lions trade the benefit of an increased chance of catching an animal for food against the size of the share for individual members of a pack when an animal is caught.
Another area where researchers are confident of deriving useful results involves the behaviour of bees. For example, swarms of bees carry out the task of gathering and storing food using two simple rules: first, nectar-storing bees unload nectar from collector bees returning to the hive at a rate that is proportional to the richness of the nectar; second, if bees are unloaded rapidly they recruit other bees to help them to gather from their food source. These two rules result in more bees collecting better nectar from richer food sources.
What is impressive about these areas of research, particularly in the insect world, is that they hold out the promise that simple rules to do with communication and task scheduling can produce groups of workers whose power is much greater than the sum of the capabilities of the individual workers. There is also the promise that large tasks we cannot even dream about now can be carried out by networks of computers and people, mediated by a small number of simple rules.
The author is professor of computer science at the Open University.