Article ID: | iaor20032833 |
Country: | United States |
Volume: | 74 |
Issue: | 1 |
Start Page Number: | 57 |
End Page Number: | 77 |
Publication Date: | Oct 2002 |
Journal: | Agricultural Systems |
Authors: | Hayman P.T., Easdown W.J. |
Keywords: | artificial intelligence: decision support |
An ecological framework is used to study the reinforcing and limiting processes on a computerised decision support system (DSS) designed for winter cropping decisions in the northeastern Australian Grains-belt (WHEATMAN). We found that WHEATMAN has had a significant impact on how many advisers structure their thinking and much of their advice on winter cropping in the region, but the number of routine users of WHEATMAN remains relatively low. Computer hardware was the most obvious limiting factor to widespread use during the early stages of the 15 year history of the project. However, despite a dramatic increase in the availability of computers on grain farms (from 5 to 75%), a maximum of 250 out of an estimated 4,500 grain farmers in the region with computers directly use WHEATMAN. Another common limiting factor for adoption of DSS is a failure to engage with end users; yet from early days the WHEATMAN project had a high degree of extension agronomist and farmer input. We suggest that just as the debate on the adoption of DSS was dominated by discussions of computerisation in the late 1980s, notions of user involvement have dominated current debate. Experiences with WHEATMAN suggest that well designed software and a focused development team approach, good access to hardware and representative end user involvement are necessary requirements to help explain the comparative longevity of the project. On their own these are not sufficient requirements for widespread adoption or impact. We argue that the perception of farmers of the nature of dryland farm management in general, and the specific decisions addressed by WHEATMAN are the primary limitations to the routine use of a computerised DSS for tactical decision making.