Simulation for Farm SuccessMonday, Apr 21, 2003
Clive Dalton (see news brief below) puts the contention that “science has farmers making the same old mistakes”. To illustrate his point he exampled the push for hogget lambing. The same push he had been part of in the 1970’s. Bonus lambs it was claimed – but at what expense to the balance of the farm business?
Farming is a system and as useful as partial budgets and limited stock performance trials may be they have the propensity when presented by a competent scientist to be accepted without question. Often the continuing total real profit being made by the individual farmer is ignored as they go about allocating feed to their priority stock in accord with the recipe. Experience shows that all stock will have a claim to priority feeding, otherwise production efficiency, performance and profitability will rapidly decline for these less favoured groups at a rate that may exceed the production increase of the newly favoured group.
The key point in the Dalton article points to the need for computer simulation of the total farming system. The ability to simulate and model farming systems has been available for some 25 years but as an extension tool it has never been used to anywhere near its potential
The reason for this may be that farm production modelling is seen a ‘black art” that, as if by magic, the future of the farm business can be foretold by some guru acting as a high priest from some obscure order visiting only high payers and the worthy. However reality is far removed from such a picture
Production modelling and computer simulation is used extensively by most production industries. As a technique it is well understood and its form ranges from simple spreadsheet models to complex LP simulations. The main barrier to farm production modelling is getting at the essence of the particular farm being modelled to enable output to be actionable management information based on the minimum of inputs.
However, having produced the model answers the next barrier is getting farmers to believe that what the model is telling them after applying the farmers own farm specific variables is useful management information.
The story is told of the smart computer modelling system developed by Team New Zealand for the Santiago competition. The system worked remarkably well displaying the strains and loadings being born by the key points on the boat and instantly advising new settings to maximise boat speed in the current conditions. The interesting outcome was that the crew felt they had a better feel for boats capabilities than that which was shown on the screen and so from time to time place double the agreed maximum loading on the forestay.
The boat story illustrates the conflict between information that is modelled and information felt through the seat of the pants or from “experience”. Modelling maximises performance within the constraints, “experience’ invariably tests the limits by finding where it breaks. And like Team NZ 2003 farmers may find they don’t get a second chance when their structure “breaks”.