The First Workshop on Blackboard Systems

The first panel was primarily concerned with identifying the architectural and organizational attributes that distinguish blackboard systems. As more and more systems appear claiming to use the blackboard paradigm, it is increasingly necessary to determine a set of criteria that define the essential elements of blackboard systems. This panel was also concerned with the question of how far systems can deviate from the basic blackboard architecture and still rightly be classified as a blackboard system. Penny Nii moderated this panel. The consensus was that the essential components of a blackboard system are the blackboard database and the knowledge sources. The blackboard database is the repository of information, and the knowledge sources represent the procedural knowledge that uses and manipulates the blackboard data structures. Also inherent in a blackboard architecture is the knowledge-oriented selection mechanism, or metalevel control, for choosing the next knowledge source to be executed. These criteria can be used to determine whether a system can indeed be called a blackboard system. However, differences in implementation, especially where control mechanisms are concerned, are to be expected because of the influence of special problem-solving requirements of particular applications. Appropriately, this panel began with Robert Engelmore’s presentation of the history and evolution of blackhe First Workshop on Blackboard Systems, sponsored jointly by the American Association for Artificial Intelligence (AAAI) and the Boeing Advanced Technology Center, was held 13 July 1987 during AAAI-87 in Seattle. A group of researchers from industry and academia gathered at the University of Washington to discuss issues and research directions in blackboard systems. It was the first blackboard workshop open for participation to the general AI audience. Since its inception with the Hearsay-II project in the early 1970s, the blackboard architecture has undergone major changes, as discussed in Nii (1987). Because interest in this technology is growing rapidly and blackboard systems differ significantly in their implementation, the organizers determined a need for a workshop in order to improve communication between the various researchers in this field, achieve a consensus on the defining characteristics of blackboards, and describe ongoing research. With these motivations in mind, the workshop was organized as a series of five panel discussions, each concerned with one of the most relevant areas of interest: (1) blackboard architecture and organization; (2) control issues; (3) parallelism, concurrency, and distributed systems; (4) performance and real-time issues; and (5) development environments. The format of each panel was determined by the moderator. In general, the moderator laid out the issues pertaining to the area and encouraged the panel members to present their perspectives. The panel members discussed the issues in the context of their own work. Their presentations were followed by discussion with the The emergence of the blackboard architecture as a widely used paradigm for problem solving led us and other members of the blackboard research community to organize a workshop. The workshop was held during the 1987 American Association for Artificial Intelligence Conference in Seattle. The main purpose of the workshop was to highlight the advances in blackboard architectures since the introduction of the paradigm in Hearsay-II and identify issues relevant to future blackboard system research. This article describes the issues raised and the discussions in each of the five workshop panels.