The blackboard system is an architectural model used in AI that facilitates collaboration among disparate, specialized knowledge sources or agents, known as "knowledge sources," to solve complex problems. By utilizing a central repository or "blackboard," each specialized module writes partial solutions or hypotheses onto this shared platform, while also reading the contributions of others to progressively build towards a complete solution. This methodology is particularly significant in domains requiring diverse expertise and has been applied in areas such as speech recognition, automated planning, and constraint satisfaction problems. The system’s structure endorses an opportunistic problem-solving paradigm, where various components can function without stringent control and contribute asynchronously, optimizing the collective output based on their specific capabilities.
The blackboard system concept originated in the HEARSAY-II speech understanding system in the 1970s and gained prominence throughout the 1980s as an effective model for collaborative problem-solving in AI. Its rise in popularity coincided with increased computational capabilities and the growing need for flexible, modular AI systems.
Key contributors to the development of the blackboard system include Raj Reddy and his team at Carnegie Mellon University, who laid foundational principles through the HEARSAY-II project. Additionally, other significant contributors include Nii and Engelmore, who further refined the approach and disseminated its practical implications across various AI applications.