EXPERT-SYSTEMS IN AGRICULTURE - STOCK-TAK ING AND PERSPECTIVES

The paper shows applications of expert systems in agriculture and, eve n more important, fields of future development. 81 expert systems have been categorized by a scheme developed especially for that purpose. T he discription of the expert systems shows that some of the systems ha ve got to the state implementation in a technical sense. They are able to support potential users in such problem areas where problem solvin g can be carried out by means of decision trees. For other problems wh ich were less structured, it is much more difficult do develop expert systems. The most sophisticated expert systems in a sence of technical ripeness, can be found in a discipline of plant-pathologie. Due to th e fact that most expert systems are less user friendly or not ready to be used yet, this technology has no impact on agriculture so far. Int erviews with experts in that area showed that this is likely to change in the future. A survey conducted with commercial software-houses sho wed that most of them are not willing to develop and sell expert syste ms because they see no market for them. The bigger software-houses hav e a different view They think that there will be a market in the futur e, but they state that they do not have the ressources to develop thei r own expert systems. Taking this into account, the authors suggest a corporation between universities and software-houses. The universities which have the 'know-how', are asked to develop the systems which can be distributed by software-houses. Areas for further development and potential economical solutions are all disciplines, which need an expe rt for solving problems. The field of strategic planning seems to have the biggest potential for that. With regard to the technical aspects of expert systems, knowledge acquisition and knowledge representation are lacking most. However, despite of all that, the paper outlines man y promising areas io;future development of expert systems.