Process and Human Factors Engineering
Research and Technology 2002
AI Techniques for Payload and Vehicle Processing Scheduling

Preparing vehicles and payloads for launch is an extremely complex process involving thousands of operations for each mission. Each operation requires a number of resources (facilities, equipment, personnel). Since several missions are in preparation simultaneously, they all compete for scarce resources. Furthermore, since many of these resources are extremely expensive and limited in number (often operating at or beyond their capacity), optimal assignment and efficient use is critically important. There are also a number of additional constraints imposed by ground rules, safety requirements, and the unique needs of processing vehicles and payloads destined for space. These challenges are compounded by the endless changes to the schedule caused by late deliveries, delayed flights, and malfunctioning equipment. To resolve the many conflicts and predict possible problem areas, operators must use a number of rules of thumb specifying where things should happen, whether they will happen on time, and whether the requested resources are actually necessary.

Although there are a number of commercially available scheduling systems, the degree of domain knowledge required for decisions and the unusual set of constraints make these of limited use. Stottler Henke Associates, Inc. (SHAI) solved these problems by applying a combination of artificial intelligence (AI) techniques to produce Aurora, a system capable of rapidly completing a near-optimal schedule. Aurora is unusual in that it combines sophisticated scheduling mechanisms with domain knowledge and case-based expert conflict resolution techniques to solve the scheduling problem. It also takes into account a number of problems unique to KSC, such as the needs to schedule floorspace and maintain certain spatial relationships among the tasks and components, in order to obtain high-quality results.



Aurora then displays resource usage, floorspace usage, and the spatial relationships among different activities graphically. Scheduling experts can interactively modify and update the schedule and can request more information about specific scheduling decisions. This allows them to supply additional information or verify the system’s decisions and override them to resolve conflicts, if necessary. In this last case, the program observes how the user resolves a given conflict and uses that cached knowledge in future scheduling to resolve similar conflicts. The interface uses comparable case-based reasoning techniques to automate repetitive tasks.

Key accomplishments and milestones:

  • June 2001: Completed proof-of-concept prototype.
  • April 2002: Released a scheduling system that provides support for spatial requirements and constraints as well as traditional scheduling needs.
  • August 2002: Augmented Aurora with a sophisticated timeframe selection system based on resource load. Developed a case-based expert kernel to support less experienced users and automate repetitive tasks.

Contact: S. Bigos (, UB-C4, (321) 867-6166
Participating Organization: Stottler Henke Associates, Inc. (R. Stottler)

Aurora Flow Editing and Display WindowAurora Flow Editing and Display Window

Figure 1. Aurora Flow Editing and Display Window

Floorspace Resource Allocation Display and EditorFloorspace Resource Allocation Display and Editor

Figure 2. Floorspace Resource Allocation Display and Editor

Biological Sciences
Range Technologies
Spaceport Structures and Materials
Fluid System Technologies
Process and Human Factors Eng