In accordance with the EU directives concerning opening-up the public transport sector to free competition, the Italian government has arranged a sectoral reform that regulates such opening-up and transfers funds and decision-making competence to the regional authorities. One of the objectives underlying the reform is to obtain, for each areal system served by public transport, a ratio between fare revenues and service costs which is greater than a set value. The reform's application and achievable results are currently being widely debated in Italy. In this context the Minister for Italian Research has funded, for a working group consisting of six Italian university departments, a research project whose aim is to produce "Guidelines for Local Transport Service Planning". One of the research lines concerns the need to define and test a system of models which might allow regional authorities to forecast the results of policies that they want to implement. The general process for developing a transit plan has been subdivided into the following, variously connected, phases or macromodels: demand; supply; modal, time and fare integration; supply-demand interaction; revenues and costs. The schematic structure of the system of models is reported. In the end, the overall predictive process may not produce satisfactory results at the first iteration. It is possible to intervene by changing the various elements of the process, formulating different policies, and hence arriving at new results. The process is reiterated until predictive results deemed valid are obtained. Each macromodel consists of different elements. The main elements are as follows: main models and algorithms, which form the engine of each macromodel; internal input data from other system macromodels; internal output data, used for other system macromodels or which make up the final results required to assess the plan; input data which are exogenous to the overall system, required for the main models; input data required for the main engines; these data may be supplied empirically or generated using subsidiary models and algorithms (what to model, expert panel, ); the data are endogenous in relation to the overall iteration of the system of models. Each predictive phase must also be able to operate at least on two different time scales: tactical and operative (short term) and tactical and strategic (medium term). In this context, each phase contains models that consider only spatial relations or models that consider space-time relations. Hence all the other elements of each phase must be arranged so as to give the input for the different predictive scales. For the covering abstract see ITRD E124693.
Samenvatting