Information Technology Reference
In-Depth Information
information can be acquired by means of active information gathering. Existing contin-
ual planning systems can deal with incomplete information. However, they usually rely
on the assumption that all possible states of a domain are known. This makes it, for ex-
ample, difficult to deal with a priori unknown object instances. Another important issue
that is not directly considered by previous work is the fact that a knowledge acquisition
task
task
1
can—like any other task—make the execution of an additional knowledge
acquisition task
task
2
necessary which might require the execution of the knowledge
acquisition task
task
3
and so on. Consider, for example, a situation where a robot is in-
structed to deliver Bob's mug into Bob's office. Moreover, let us assume that the robot
does know that Bob's mug is in the kitchen, but does not know the exact location of
the mug. In this situation, the robot needs to perform a knowledge acquisition task that
determines the exact location of Bob's mug. However, in order to do that via percep-
tion the robot first needs to go into the kitchen. If the robot does not have all necessary
information in order to plan how to get into the kitchen (e.g., it is unknown whether
the kitchen door is open or closed), then it needs to first perform additional knowledge
acquisition tasks that acquire this information. Existing continual planning approaches
usually fail to cope with such a situation. In contrast, we propose a continual planning
and acting approach that is able to deal with these kind of situations and thus can enable
an agent to perform tasks in a larger set of situations.
We are trying to give an answer to the following questions: How can an agent
determine knowledge acquisition activities that make it possible to find a plan when
necessary information is missing? When is it more reasonable to acquire additional
information prior to continuing the planning process? How to automatically switch be-
tween planning and acting?
The main contributions of this work are:
-
to propose the new HTN planning system
ACogPlan
that additionally considers
planning alternatives that are possible with respect to a consistent extension of the
domain model at hand, and is able to autonomously decide when it is more reason-
able to acquire additional information prior to continuing the planning process;
-
to propose the ACogPlan-based, high-level control system
ACogControl
that en-
ables an agent to perform tasks in open-ended domains;
-
and to present a set of experiments that demonstrate the performance characteristics
of the overall approach.
2
HTN Planning in Open-Ended Domains
In this section, we present the continual HTN planning system ACogPlan. We describe
the planning phase of the overall plan-based control system.
2.1 General Idea
The proposed planning system ACogPlan can be seen as an extension of the SHOP
[10] forward search (i.e., forward decomposition)
Hierarchical Task Network (HTN)
planning system. The SHOP algorithm generates plans by successively choosing an in-
stance of a
relevant
1
HTN method or planning operator for which an instance of the
1
As defined in [4, Definition 11.4].
