Engineering topics
Publications and source records attributed to Lansky, Amy L..
This paper discusses the use of a domain-independent planner, COLLAGE, as a software assistant to Earth scientists working with remotely-sensed and ground based data sets. The planner can be viewed as an advisory agent that helps scientists select appropriate data and creates a suitable plan for data-processing that meets stated scientific requirements.
This paper presents an approach to domain representation and planning that is fundamentally different from traditional methods; an approach based strictly on actions and their interrelationships, rather than on state-based goals and preconditions. In particular, we focus on the action-based planner COLLAGE, describe its methods for plan-construction, and contrast them with more traditional approaches to planning. Experiences with COLLAGE in realistic domains have shown that the action-based approach is not only more natural to use, but can also be more cost-efficient than traditional planning methods.
Localization is a general-purpose representational technique for partitioning problem requirements into subproblems. Localized search utilizes a partitioning by breaking the overall search space into several smaller search spaces, one for each subproblem. Unlike most methods of partitioning, localization allows for subproblems that overlap -- i.e., multiple search spaces may be involved in constructing shared pieces of the overall plan. In this paper we focus on two criteria for forming localizations: abstraction and scope. Abstraction-based localizations are motivated by activation relationships between problem requirements. In contrast, scope-based localizations are motivated by the relevance of problem requirements to specific portions of the overall plan. This paper also describes a method for automatically generating scope and abstraction-based localizations. Empirical results are provided that contrast their use in an office-building construction planning domain.
COLLAGE is a domain-independent planner that differs from traditional planners in two key ways: (1) it uses a diverse and extendible set of (it action-based) plan constructions methods; and (2) it utilizes a technique called (it localization) to partition the planning problem into smaller (and potentially interacting) sub problems. In this paper, we describe these features and how they are exploited in two different real-world planning domains. The first domain, building construction planning, has served as a framework for us to investigate how localization can improve scalability. The second domain, data analysis planning, benefits from COLLAGE's unique plan construction and representation methods. In addition, we describe extensions that were necessary to tackle these real-world problems.
There has been much recent work on the use of abstraction to improve planning behavior and cost. Another technique for dealing with the inherently explosive cost of planning is localization. This paper compares the relative strengths of localization and abstraction in reducing planning search cost. In particular, localization is shown to subsume abstraction. Localization techniques can model the various methods of abstraction that have been used, but also provide a much more flexible framework, with a broader range of benefits.
Much of commonsense knowledge about the real world is in the form of procedures or sequences of actions for achieving particular goals. In this paper, a formalism is presented for representing such knowledge using the notion of process. A declarative semantics for the representation is given, which allows a user to state facts about the effects of doing things in the problem domain of interest. An operational semantics is also provided, which shows how this knowledge can be used to achieve particular goals or to form intentions regarding their achievement. Given both semantics, the formalism additionally serves as an executable specification language suitable for constructing complex systems. A system based on this formalism is described, and examples involving control of an autonomous robot and fault diagnosis for NASA's Space Shuttle are provided.
A high level of automation is of paramount importance in most space operations. It is critical for unmanned missions and greatly increases the effectiveness of manned missions. However, although many functions can be automated by using advanced engineering techniques, others require complex reasoning, sensing, and manipulatory capabilities that go beyond this technology. Automation of fault diagnosis and malfunction handling is a case in point. The military have long been interested in this problem, and have developed automatic test equipment to aid in the maintenance of complex military hardware. These systems are all based on conventional software and engineering techniques. However, the effectiveness of such test equipment is severely limited. The equipment is inflexible and unresponsive to the skill level of the technicians using it. The diagnostic procedures cannot be matched to the exigencies of the current situation nor can they cope with reconfiguration or modification of the items under test. The diagnosis cannot be guided by useful advice from technicians and, when a fault cannot be isolated, no explanation is given as to the cause of failure. Because these systems perform a prescribed sequence of tests, they cannot utilize knowledge of a particular situation to focus attention on more likely trouble spots. Consequently, real-time performance is highly unsatisfactory. Furthermore, the cost of developing test software is substantial and time to maturation is excessive. Significant advances in artificial intelligence (AI) have recently led to the development of powerful and flexible reasoning systems, known as expert or knowledge-based systems. We have devised a powerful and theoretically sound scheme for representing and reasoning about procedural knowledge.