Construction is a complex undertaking, which can be interpretated from three different perspectives. Koskela (2000) explores the Process perspective of construction in depth, but this perspective mainly. However, in his dealing with the peculiarities of construction he identifies two other perspectives deserving a similar in depth study: the nature of the Project and the nature of the Industry. Only by combining all three perspectives can a whole picture of construction be made.

This letter does not try to explore the two other perspectives in great depth, it outlines the landscape in general terms only.

It focuses on the peculiarities of each of the three perspectives and it uses these peculiarities to understand construction as a complex phenomenon.

Looking at construction from the process perspective one find a process characterized by:

These seven prerequisites are:

In this context, the nature of the prerequisites is not the issue, just their existence. They represent seven flows penetrating the project and the industry. One activity can thus be seen as a node in highly connected feed forward network. Once an activity's inflows are fulfilled, the activity 'fires' – to use a phrase from the neural network theory – when completed by releasing seven resources flowing on to new activities. These activities may be within the same project – as are normally the case for previous activities and space, sometimes the case for information, labor and equipment, whereas it for external conditions and materials may just as well be to other projects.

O'Brien (2000) as a matter of fact studies the management of the flow of labor through different projects from the contractor's viewpoint.

The distance between design and production stems partly from the artistic nature of architecture, coining the architect an artist. But the division of work within the industry as a whole enforces it. The division by trades – ie by skills – is more common than a division by products as seen in the manufacturing industry.

The separation between design and production is probably the main reason for concurrent engineering not being the normal practice within the industry. It may also be the reason for not understanding construction as a process in the way the manufacturing industry does.

The general viewpoint is that the production management in construction is the responsibility of the project management, but this is not the case. The project management must – enforced by the nature of the project – manage the contracts with the industry, i.e. the participating companies. It can thus not engage itself in the actual production or in its management.

As the industry – up till Koskela (2000) – has not possessed a formal understanding of its own undertakings, this fact has been obscured.

Construction makes its production through projects. Projects are characterized by:

The one of a kind nature of the project adds to the complexity. One-of-a-kind undertakings are always experiments or prototypes in some aspects. This causes unforeseen events, which again give rise to improvisation. All this decreases order and increases chaos.

The on site production with its temporary production facilities adds further to the complexity. Again improvisation is the name of the game. Indeed, looking at practice gives one the impression that project management is valued mainly by its capacity for improvisation.

The temporary organization is also a source for complexity. Not only by its nature, where the participants are strangers to each other, but also by the lack of capacity for common learning from experience.

Once their part of the job is completed, the parties are up and away for their next task.

Looking at the industry one finds a number of peculiarities too:

These peculiarities interact, and they can all be understood as consequences of the division of work. But they add to the complexity as well.

The nature of construction is that of a service industry. The companies offer skills, not products. This causes a low threshold for entry. Indeed, any skilled craftsman can enter without any substantial capital, making use of the tools he already possesses. This again causes fragmentation, with most firms operating locally only.

High fragmen tation causes low earning rates, which again give little capitalization. The competition mainly based on cost – as opposite to value – adds to this phenomenon, which again gives rise to the little or nonexistent innovation.

The industry nature calls for a lot of detailed regulation setting rules to protect the small agents, the nature of the industry etc. This regulation reflects a number of interests, all concerned with other issues than the industry's productivity, and it adds strongly to its continuing being highly complex.

Koskela (2000) proposes the use of a Transformation-Flow-Value model in the understanding of construction. Indeed, this model explains a number of peculiarities, and it is a useful tool for rethinking the construction process.

But T-F-V dimensions can also be seen as the characteristics of the three perspectives in the Process-Project-Industry model introduced above.

The Industry perspective deals with the T-dimension mainly, at least for the time being. The participants in the industry look upon their undertakings as transformations and they deal only reluctantly with the flow and value aspects.

The Process perspective treats the flow-dimension as seen in the present lean development. And this will probably be the case as long as construction works through projects mainly.

The Project perspective views the V-dimension. The value generation takes mainly place through the design process' generation of the genotype of the constructed artifact, which is realized through the construction phase's production of the phenotype.

The complexity of this system is further emphasized by the management structure. The project is managed by the project management, but the involved agents are each managed by their own business management. And the common undertaking of the highly complex, non-linear, dynamic – and thus near chaotic – construction process is not managed at all.

The Industry model was developed during the summer 2000. It was made as a basis for the preparation of project proposals for the ongoing Danish program for the development of the building industry: Project House.

In this, the complex nature of the building process was an important understanding, explaining a number of the industry's peculiarities and guiding the strategy of the project proposals.

It is the author's intention to further use the model for analyzing the strategies applied in the recently completed Danish program: Process and Product development in the Building industry – PPB.

However, it is also the author's opinion that the model deserves a more in depth study on a formal basis, which hopefully can be made as part of the Lean Construction Initiative.

Ballard, Glenn (2000): The Last Planner System of Production Control, Scool of Civil Engineering, Faculcy of Engineering, The University of Birmingham

Koskela. Lauri (1999): Management of Production in Construction: A Theoretical View IGLC-7

Koskela, Lauri (2000): An exploration towards a production theory and its application to construction, VVT Technical Research Center of Finland

O'Brien, William J. (2000), Multi-Project Resource Allocation: Parametric Models and Managerial Implications, IGLC-8, Brighton 2000