RegMAS borrows many concept from AgriPoliS, but it also progress on many areas:
Spacial dimension is explicit and initialized from real land-use data, using satellite informations. Plots are explicitly modeled in the MIP as individual resources, while in AgriPoliS they are aggregated according to their land type (e.g. under the resource “arable land”). Aside the opportunity of seeing where production happen, this implementation leads to easy improvements of the model over other layers of spacial information. E.g. it is relatively easy to insert an information about altimetry that influence individual plots revenue capacity.
The share of borrowed capital for acquire investments is not “fixed” with the investment object, but farmers are free to chose case by case the amount that minimize the costs;
Farmers have greater sensibility of the overall environment, including extra-agricultural variables;
The model is designed from the ground up to explicitly consider farmers as one type of several possible type of agents;
The model code is
publicly available and free from charges, open to review and experiments are easily repeatable;
The program use state-of-the art libraries for multi-platform porting and endorse only free software tools.
On the other hand AgriPoliS has a long list of features that enhance the capability of the model of catching real world phenomena that are not scheduled on present in RegMAS. Among them, the technological improvement of new investments over old investments, the increase in opportunity costs during farm generational shift, and the differentiation between short-term and long-term interest rate.
