The complexity of and highly capital intensive modern agricultural producti
on systems stimulates the importance of operational management within agric
ultural mechanisation systems. The development of model based decision supp
ort systems related to the planning and execution of farm operations is ess
ential considering a sustainable and profitable production. The planning mo
dels generally applied in both industry and agriculture have been tradition
al deterministic operational research models trying to provide the planner
with exact optimal solutions. However, often these models have not been abl
e to capture the inherent uncertainty and risks involved when planning futu
re actions, and thereby prompting the need for models exceeding the limitat
ions of traditional modelling approaches.
In an attempt to provide for a more comprehensive modelling of planning as
part of a decision support system, the use of probabilistic networks (Bayes
ian networks) has been investigated as a method for building planning model
s for scheduling field operations. The Bayesian approach includes principle
s from Artificial Intelligence and provides for a normative framework to mo
del decision making and reasoning under uncertainty. For instance, in the c
ase of managing harvesting operations, the model will give the probable num
ber of harvesting hours, probable state of the crop with regard to quantity
and quality, probable machinery capacity, when evidence on area to be harv
ested, weather prognosis, etc. are, provided.