A multi-disciplinary research approach is demanded to improve our understanding about corporate memories and their use, how to develop them, how to manage them, and how to integrate them into the on-going practice of a company.
The following two research questions (on group work and on concurrent
engineering) are motivated by the aim to achieve the first and third
use of corporate memory at the macro perspective
(page 
), i.e. enable project-oriented rather than
functional-oriented work and enable virtual corporations.
Group work: How does a heterogeneous group work together in practice? What are the enabling and stimulating conditions (at the information and social level) to arrive at a cooperating group and at group decisions? How can these conditions be supported and encouraged?
An example of a condition stimulating cooperation is a global goal that is common to all team members. Contributions of each team member are evaluated only with respect to this common groups goal. For example, in an engineering project, the global goal may be customer-satisfaction. The team as a whole is evaluated with respect to that criteria. An individual is evaluated with respect to the value (s)he added to that criteria. This criteria is not the same as, say, the degree the engineered device matches the initial device specification because the latter does not necessarily imply an improved customer satisfaction (because what if the document does not well reflect the customer's needs or these needs have evolved?). Moreover, this latter criteria naturally causes a tension between a team member who wants to change the device spec. and the engineer who wants to keep the device spec. stable. This is counterproductive to team work. What are some other conditions? A study of findings in Social Sciences (e.g., on social dilemmas), Organizational Sciences (e.g., how do decisions happen in organizations?), Groupware, and CSCW seems to be relevant for this purpose.
Concurrent engineering: How to discover early in the design process the requirements that are pertinent to the different life-cycle phases of a device? How to consider these requirements concurrently during the design process? How can the different life-cycle analyses and development efforts be integrated in a manner that one serves the other?
The next research question is motivated to enable corporate learning
(second use of corporate memory at the macro perspective,
page ).
Enable Corporate Learning: How can we enable and stimulate corporate learning to improve and enhance its key know-how? How to support creativity? Dynamics? Motivation?
This, in my opinion, seems to be one of the most crucial research questions on corporate memory. If there is to be a large practical interest in this topic, we need to understand how know-how available in a company can be exploited in new ways that enable the company to grow their market share, to conquer new markets (i.e. to realize projects that it couldn't realize before), to enable learning from past experiences such that work can be realized more efficiently and with a better quality. The biggest mistake a company can make is to look with great satisfaction at its know-how and to use this know-how in situations similar to the past (only). The term memory has this connotation and it may scare off industrial groups that realize that learning, creativity, anti-routine, are the key words to success.
I believe before we can start thinking about how to build a corporate memory, we need to understand the dependency between corporate memories (their internal structure, their use, its requirements, its applicability conditions) and the type of organizations in which they will be embedded.
Organizational analysis: What are the different types of enterprises and, for each type, what are their common organizational structures? How do these enterprise types and organizations affect the structure, functionality, and use of a corporate memory?
For example, one type of enterprise is an engineering company (like Dassault). These companies have the tradition of being strongly structured along functional disciplines and with a hierarchical authority and communication structure. One obvious role of a corporate memory is to support cross-functional work to enable concurrent engineering. Another role might be to bypass the bureaucratic communication delays and permission demands inherent in large hierarchical structures. A corporate memory could support a direct communication among project team members (independent of their departments) and to give local decision-making authority and control (e.g., through centralized information supply) to team members to quickly respond to customer needs.
Alternative, a service company (such as Andersen Consulting) has the tradition of being loosely structured in small dynamic teams. Different aspects of a corporate memory will have to be stressed for these type of companies and their internal organizations. Their market is a lot more dynamic than that one of an engineering company and therefore, the corporate memory should support efficiency in the way these teams are setup and the way they work. For example, because Andersen is a world-wide consulting company, the importance of identifying and consulting relevant sources of expertise (humans) throughout their offices in the world in a quick manner is crucial. Indeed, there is often only little time to setup a project team (in particular of busy international members), to write and submit a project proposal as soon as a market opportunity is recognized. A corporate memory for such service companies should stress this issue.
There is already a lot of research available on design rationales and design history. I believe a lot of this research is useful to achieve the use of a corporate memory at the micro level.
Project history and design rationales: What (intellectual) information is to be captured about a project, its course, and the development/service process to enable (other or the same) project(s)s to perform better or to avoid a loss of know-how? How to capture this information in a manner that minimizes overhead? What is the use and benefit of this information? How is this information to be organized and managed for maintenance and (re)use?
Often, to arrive at a final design, a lot of complex decisions in uncertain situations have to be made. Design rationales capture these design decisions, their context and motivations. The essential contribution of design rationales is that it records intellectual information that would have been lost (or remained in an individual's head) after some time, that has to be continually re-invented or uncovered.
Models of individual expertise: What information about individual experiences needs to be captured and stored? How can it evolve? Are models of expertise pertinent to a Corporate Memory? What are different levels of detail that this expertise can be described to be useful? How can these descriptions be used?
WWW and software agents: How can the metaphor of the World-Wide-Web be used to organize and manage a corporate memory? How can the idea of software agents be explored to add an active and dynamic dimension to the memory concept? What other technologies are useful or needed to realize a corporate memory?