Our research is developing persistent agents that can perform complex tasks
in dynamic and uncertain environments. We refer to such agents as
taskable, reactive agents. An agent of this type requires a number
of capabilities. The ability to execute complex tasks necessitates the use
of strategic plans that provide an outline for accomplishing tasks; hence,
the agent must be able to synthesize new plans at run-time. The dynamic
nature of the environment requires that the agent be able to deal with
unpredictable changes in its world. As such, agents must be able to react
to unanticipated events by taking appropriate actions in a timely manner,
while continuing activities that support current goals. The
unpredictability of the world could lead to plan failures. Thus, agents
must have the ability to recover from failures by adapting their activities
to the new situation, or replanning when the world changes sufficiently.
Finally, the agent should be able to perform all of the above operations
even in the face of uncertainty about the current and future states of the
Cypress is a system for defining taskable, reactive agents. At its core
are mature, powerful planning and execution technologies, namely the SIPE-2 generative planner and the PRS-CL reactive execution system. Since PRS-CL and
SIPE-2 employ different internal representations for plans and actions,
Cypress supports the use of an interlingua, called the Act
formalism, that enables these two systems to share procedural knowledge.
Using the Act interlingua, PRS-CL can execute plans produced by SIPE-2 and
can invoke SIPE-2 in situations where run-time replanning is required.
The Act-Editor provides a means of graphically
viewing and editing this procedural knowledge. A similar capability for
declarative knowledge is provided by the GKB-Editor.
Together these support ready access to a common store of knowledge by all
Cypress componenets, as well as, the Cypress user.
Gister-CL implements a suite of evidential
reasoning techniques that can be used during plan generation and plan
execution to analyze uncertain information about the world and possible
actions. For example, Gister-CL can be used to reason about prevailing
conditions in order to choose the most effective Act among several
candidates, or to verify the state of the world based upon interpreted
Cypress' components operate asynchronously, in a loosely-coupled
fashion. This makes it possible for the component systems to run in
parallel, even on different machines, without interfering with the actions
of the others. While the subsystems of Cypress can function independently,
Cypress is used most advantageously as an integrated framework. This is
best illustrated through an example.
D. E. Wilkins, K. L. Myers, J. D. Lowrance, and L. P. Wesley, "Planning and Reacting
in Uncertain and Dynamic Environments,"Journal of Experimental and
Theoretical AI, vol. 7, no. 1, pp. 197-227, 1995.
D. E. Wilkins and K. L. Myers, "A Common Knowledge
Representation for Plan Generation and Reactive Execution," Journal of
Logic and Computation, in press.
D. E. Wilkins, K. L. Myers, and L. P. Wesley,
"Cypress: Planning and Reacting
under Uncertainity," in ARPA/Rome Laboratory Planning and Scheduling
Initiative Workshop Proceedings (M. H. Burstein, ed.),
Morgan Kaufmann Publishers Inc., San Mateo, CA, pp. 111-120, Feb. 1994.
John D. Lowrance,
Karen L. Myers,
David E. Wilkins,
Artificial Intelligence Center
Act-Editor, Cypress, Gister, Gister-CL, GKB-Editor,
Grasper-CL, PRS-CL, SIPE, and SIPE-2 are trademarks of SRI International.
Copyright © 1995 SRI International, 333 Ravenswood Ave., Menlo Park, CA 94025 USA.
All rights reserved.
David E. Wilkins email@example.com
Last modified: Thu Nov 11 20:09:14 1999