In a conventional highway design model, the data consist of points and lines that define the outline of the planned works. Other information critical to successful completion such as design data and specifications is disassociated. In typical engineering application data is not only stored in different locations, it is often linked via a human or paper intermediary. This segregation of data occurs because, in the common file based data management environment, there is no direct link between a drawing and a document. Changes in one may not be reflected in the other. This state of affairs is accepted historically as the design process has evolved to produce a final paper output with documents and drawings as separate entities. This is the focus of established design processes that have evolved over the years.
In a typical large highways project under the Early Contractor Involvement model, systems such as Business Collaborator (BC) share information among the parties. However in a typical application the data set held on BC is incomplete. BC is often used to issue information to team members rather than as an information or knowledge repository. The effect of this approach is to create duplication between data stored in the collaboration system and data held internally on file servers. This duplication may further increase inside the design organisation as different teams or groups maintain their own file systems. Multiple asynchronous copies of information may be held (and relied upon) that are not subject to a consistent update policy.
The lack of compatibility means that it is not possible to go to a “single source of truth” for project information it is necessary in many cases to know first what is being looked for and where it may be stored. This increases the workload to locate information and introduces a risk that it may not be found or when found may be (dangerously) out of date. A user cannot be totally confident that he has discovered the true and complete information needed to complete his task. The secondary effect of this is that often it will be easier for a person to go direct to the source of the information and the result is uncontrolled information. Additional workload is created within teams from making and responding to these requests. Multiple requests may be made over time for essentially the same information.
Should information need updating it is not possible to be confident that all versions on the system have been updated, and team members, as a result, have low confidence that information on the system is up to date. The output may, therefore, be subject to additional and unnecessary validation and control stages. Authors may hold back information in which they have little confidence, but is still useful information. This creates additional inefficiencies.
Where out of date information is held on the system this may result in rework when new information comes to light or it is corrected. A culture of making do emerges where tasks are commenced even though faith in the information is reduced, or it is known that input data will be subject to change. To enable work to advance assumptions are made to avoid future correction that in themselves may require correction at a later date. Hidden contingency is built in to cover these assumptions increasing base costs and extending project time.
Within many organisations the culture of “design iteration” remains embedded as it is implicitly accepted that information will change through the design cycle and that the design will be changed several times in response to this changing information. Supported by better information and better management of that information it should be possible to deliver better designs, earlier and cheaper, by eliminating wasteful iteration, risk and rework.
To adopt new ways of working and managing information, we must first recognise that deliverable and document focused systems are based on a paper model. In such a model information is collated into paper based documents such as reports and drawings. Information based methodologies eclipse paper and focus on delivering the right information to the right person at the right time. How information is delivered, and consumed can be variable, depending on the receiver. It is not necessary, therefore, to maintain information in different forms just because the consumer of that information has different needs. Information kept in different forms may result in omissions when updated if one of the forms is overlooked.
Much current technology for managing and indexing information is based on the assumption that information arrives in paper form or a paper analogue (e.g. email). In fact, what is essential is that the information can be found by those needing to know it. Such information is better found, not from the paper mindset of looking through all likely documents, but through search.
Traditional filing systems are based on a paper methodology with information collated into folders (files) browsed as a file analogue. The division into files can be arbitrary and at odds with how information is sought by the user. Within a traditional filing system, vertical navigation is straightforward, but horizontal navigation between folders is more difficult. The analogy would be that to move from the east wing of a 40 floor building to the west wing it was necessary to take a lift first to the lobby. If the constraint of the paper is broken, and the appropriate technology adopted, it becomes possible to obtain information by whatever method (for example full text search) is best and most effective.
It is clear, therefore, that paper based approaches to information management in a world in which almost all project information is either electronic or convertible to electronic form creates an unnecessary overhead. Search based approaches will locate information more thoroughly and more efficiently. We are used now to the internet and Google, would we gladly give up Google to use a library instead? By continuing to use deliverable and document focused systems, as opposed to information focused, this is in effect what we force teams to do.
An effective information system must provide assurance as to the quality of the information being delivered. The current position of document controller on a project will change to information manager in order that the system is properly managed and controlled.