Integration of Problem-Solving Techniques in Agriculture

Accessibility for Conventional Applications Integrating KBSs with simulations, databases, and other conventional applications increases the audience for these applications. Several simulations developed within academic and research organizations have not been used to their potential because they are difficult to apply. Even when such models and simulations are used, the danger exists of their being used inappropriately because of insufficient user knowledge. In some cases, KBSs have been integrated as intelligent front ends, or data collectors, into applications to (1) remove the burden of syntax and format and (2) apply domain-specific knowledge to request user information expressly required by the integrated application. In other cases, such systems can provide best guesses when the user is unable to provide input. KBSs have also been used as output filters, or data interpreters. A continuing need exists for the development of such back-end systems for simulations, optimization programs, and griculture is an area of enormous potential for applications involving integrated KBS and conventional technologies. For example, excellent databases are available for information ranging from historical weather data to individual dairy cow records. Complex simulations have been developed to describe phenomena from plant growth to economic systems. Such investments are valuable as knowledge sources for knowledge-based decision making. The primary goals of the workshop were to (1) assess the state of the art of integrated systems for agriculture, (2) determine factors necessary to advance the state of the art, (3) expose research needs and opportunities for the future, and (4) form an interdisciplinary core of researchers for future communication and collaboration. The workshop was organized to ensure that these goals were reached. During the first two days, papers were presented by a selected group of participants to set the stage for discussion and help define the state of the art. Presenters had approximately 45 minutes to describe the philosophy of their approach and the implementation details. Several completed systems were on display during discussion and break times. Breaks were scheduled after every other presentation to provide time for informal discussions. It was during these free periods that the most fruitful interaction took place. The third day was devoted to discussing and summarizing the issues raised throughout the workshop and to plan future directions. Problem-solving techniques such as modeling, simulation, optimization, and network analysis have been used extensively to help agricultural scientists and practitioners understand and control biological systems. By their nature, most of these systems are difficult to quantitatively define. Many of the models and simulations that have been developed lack a user interface which enables people other than the developer to use them. As a result, several scientists are integrating knowledgebased–system (KBS) technology with conventional problem-solving techniques to increase the robustness and usability of their systems. To investigate the similarities and differences of leading scientists’ approaches, a pioneer workshop, supported by the American Association for Artificial Intelligence (AAAI) and the Knowledge Systems Area of the American Society of Agricultural Engineers, was held in San Antonio, Texas, on 10–12 August 1988. Part of the AAAI Applied Workshop Series, the meeting was intended to bring together researchers and practitioners active in applying AI concepts to agricultural problems.