Current automated planning and scheduling tools are generally single-user, largely "standalone" systems based on heterogeneous languages, development tools, and processing platforms. Domain models (e.g. of tasks, resources, and constraints) are typically incompatible across systems. Interfaces to users, databases, and other decision and operations support tools tend to be custom-built and difficult to extend. The goal of this effort is to design an innovative framework for non-intrusively integrating and coordinating intelligent and conventional systems with one another and with multiple users to support concurrent, distributed solution of large planning and scheduling problems. The proposed effort will leverage previous work on advanced, object-oriented tools for distributed communication, control, and information modeling. Uniform representational models and mapping capabilities will be designed for transparently moving data, knowledge, and control structures across disparate tools and information stores. Knowledge-based cooperative control capabilities will be applied to manage access to work elements, route tasks to suitable human or tool servers, and detect and help resolve inconsistencies across elements of "Master" schedules or plans. These advanced integration and coordination capabilities will enhance the quality and productivity of planning and scheduling activities for Shuttle operations and other NASA missions.Architectures for integrating and coordinating software systems for automating decision and operations support have commercial applicability for manufacturing process control and for communications, computer, power, and transportation networks.cooperating intelligent systems, distributed systems, planning schedulingSTATUS: Phase I Only
