The Automated Planning and Scheduling Environment (ASPEN) version 2 computer program comprises a modular, reconfigurable software framework and collection of software components that can be used for automated planning and scheduling in a variety of applications. ASPEN can automatically generate schedules pursuant to high-level goals specified by the user. ASPEN can also provide automated assistance, with human intervention, in the correction of previously generated faulty schedules. The primary advantages of ASPEN are simplicity and ease of use. Features include a heuristic specification language to provide guidance for the automatic-scheduling software components, an external function interface that facilitates integration with other software, and a graphical user interface for viewing and manipulating schedules. ASPEN was written in C++ for execution on a Sun Workstation running Solaris 2.5-2.6 with at least 32MB of random-access memory and 1GB of disk storage. The Objectspace STL Library and Java Runtime Environment are necessary for execution. The graphical user interface (GUI) can also be run on a PC with the Java Runtime Environment.
This work was done by Robert Sherwood, Steve Chien, Gregg Rabideau, Anita Govindjee, Alex Fukunaga, David Yan, and Russell Knight of Caltech for NASA's Jet Propulsion Laboratory. NPO-20299
This Brief includes a Technical Support Package (TSP).
Easy-to-use software for planning and scheduling
(reference NPO20299) is currently available for download from the TSP library.
Don't have an account?
Overview
The document discusses the Automated Planning and Scheduling Environment (ASPEN), a software framework developed by NASA's Jet Propulsion Laboratory (JPL) aimed at improving spacecraft mission planning. The primary goal of ASPEN is to reduce operational costs by automating the spacecraft mission planning process, addressing the challenge of acquiring and expressing spacecraft operations knowledge from personnel in a user-friendly manner.
ASPEN is designed for domain experts who may not have a background in automated planning technology, allowing them to construct plan models intuitively. This is achieved through a mixed-initiative planning mode and a graphical user interface (GUI) that enables operations personnel to easily edit automatically generated plans. The document highlights that a non-expert with an operations background successfully created ASPEN models for two missions, EO-1 and UFO-1.
The framework supports the representation of spacecraft operations knowledge in a way that aligns with the traditional spacecraft execution environment. It emphasizes the need for fixed execution times for activities, as traditional execution engines cannot handle time intervals. ASPEN defines various types of constraints that activities can impose on plan elements, including temporal constraints, functional dependencies, resource reservations, and state reservations.
Temporal constraints are particularly detailed, with six defined relations: starts_before, starts_after, ends_before, ends_after, contains, and contained_by. These relations help establish the timing and sequencing of activities within a mission plan. Additionally, ASPEN categorizes resources into aggregate and atomic types, with aggregate resources having finite capacities that can be allocated among multiple activities.
The document also touches on the modular design of ASPEN models, allowing for flexibility and ease of use. It emphasizes that modelers do not need to understand the underlying planning engine, making the tool accessible to a broader range of users.
In summary, the ASPEN framework represents a significant advancement in spacecraft mission planning, enabling efficient and intuitive planning processes while reducing the knowledge acquisition bottleneck. Its design caters to the needs of operations personnel, facilitating the creation of effective mission plans for various space missions.
