The main objective of the course is to teach and to train students to deal with transport planning activities through a scientific approach. The main steps of the transport modeling and planning process are presented at lesson and contextually developed by students through workshops on a real case study. Special attention is paid to those aspects that are critical for the quality of a transport plan, from data collection issues to the appropriate selection of a model for the problem instance under consideration. The last part of the course enlarges the perspective to show to students the interactions between transport systems and economics, land use and environmental issues, and how these interactions impact evaluation activities of a transport scenario.

Ability to analyze mobility data through basic statistical techniques. Ability to understand the ambits of use and limitations of the most common transport planning models and methods to forecast travel demand and its interaction with the supply. Ability to carry out the most common quantitative analyses that are normally needed to set up a rationale transport planning process. Ability to consider and to quantitatively evaluate some key aspects of the mobility phenomenon within a complex framework in which economic, territorial and environmental issues interfere with transport systems.

Basic algebra and analysis (matrices, derivatives, integrals). Fundamentals of statistics (probability distributions, cumulative distribution functions, mean, variance, standard error, coefficient of variation, confidence intervals, regression and least squares, the central limit theorem). Fundamentals of economics (demand and offer curves). Concerning informatics, advanced use of a spreadsheet, fundamentals of computer programming and eventually basic notions concerning GIS (geographic information systems).

1. Introduction and surveying phase (9 hours of lessons and 15 hours of workshops)
• Origins of the discipline. Some characteristics of travel demand: derived goods, relationship with the economic development, travel time budgets. The main stages in the transport planning process: surveying, simulation and modelling, evaluation and forecast.
• Definition of study area and enlarged study area. The notions of trip, tour, stage. Trip representation: zoning and zoning criteria. Definitions of trip origin and destination, trip productor and attractor. Representation of transport networks: concept of graph, centroid and its localisation, centroid connector, O/D matrix, arc attributes.
• Passenger travel demand surveys: kinds of surveys and questionnaire design. Simple, stratified and choice-based sampling.

2. Simulation phase: transport models (16 hours of lessons and 13 hours of workshops)
• Simulation phase: transport model definition. Concepts of specification, estimation, calibration and validation of passenger travel models. Models classifications. Kinds of errors in models. Introduction to the four-step model.
• Trip generation. Classification of trips. Aggregate trip regression models for attractions (zoning level) and disaggregate trip regression models for productions (household or individual level). Non linearity and aggregation of the results. Category analysis models. Balancing trip production and attraction models.
• Trip distribution. Recall on generalised trip costs and value of travel time. Singly and doubly constrained models. Growth factor methods. The gravity model.
• Modal split: aggregate models. Factors that influence the choice of the transport mode. Notes on aggregate modal split models and their positioning within the four-step transport planning process. Notes on synthetic trip distribution and modal trip models and on direct demand models.
• Modal split: disaggregate discrete choice models. Logistic regression and linear probability model, logit and probit model formulation. Consumer theory: concept of random utility, related assumptions and subsequent characteristics of random utility models. The multinomial logit model: statistical assumptions on the distribution of residuals and subsequent limitations of the model, independence from irrelevant alternatives. Heteroskedastic and correlated error structures: the hierarchical or nested logit model, the cross-nested logit model. The multinomial probit model. The mixed logit model: random coefficients specification.
• Using discrete choice models: model specification, choice set definition, model estimation, aggregation problems. Non-compensatory choice protocols: dominance, satisfaction and lexicographic rules; role of habit.
• Trip assignment. Relationship between costs and flows, path choice mechanism and minimum spanning tree, demand-supply equilibrium. Deterministic assignment methods: all-or-nothing. Notes on the stochastic methods of multipath assignment. Capacity constrained assignment, Wardrop principles and Braess’s paradox. Limitations of the classical trip assignment methods.
• The four-steps model critique and activity-based approaches. Model implementations: notes on the transport planning software packages that are based on trips and zoning (EMME, Visum, TransCAD, Cube, OmniTrans) and on those based on agents and microsimulation (TRANSIMS, MATSim).

3. Building transport planning scenarios: evaluation and forecast (7 hours of lessons)
• Evaluation phase of a given scenario. The dimensions of an evaluation process. An example of impacts quantification: emissions and dispersion models of pollutants. The monetization of the impacts: external and social costs, travel time value and willingness to pay.
• Forecast phase. Spatial and temporal transferability of demand and supply model parameters. Updating the model exogenous variables. Transport and land use, accessibility. Induced traffic. Travel demand and mobility management tools: regulations, pricing, offer of new services.
• The transport planning process and documents in Italy. Planning at the national level: national transport plan, SIMPT . Planning at the regional and local level: regional transport plan, urban mobility plan.

Eight workshops will be held during the course; each workshop will deal with a particular task in the transport planning process, but all workshops will be based on the same case study and data and therefore follow a sequential structure. During workshops, students will have to replicate on their own laptop what is done by the instructor. Then, they will have to work in groups to fully develop the proposed activity. Workshop homework will be assigned and must be turned in shortly after for correction. Homework that is not sufficiently developed or contains errors will be rejected and the group will have to resubmit it until it is approved. Each workshop builds on the results of the previous ones, so it is essential to stick to the deadlines.

The reference book is Ortuzar, J.d.D. and Willumsen, L.G. (2011) Modelling Transport – 4th edition, Wiley, ISBN 978-0-470-76039-0, although the course covers less than half of its contents. A lot of material from other sources is needed and will be distributed by the instructor through the course webpage.

As a precondition to take the exam, homework from all eight workshops must be previously approved by the instructor. The exam is an individual oral colloquium of about 45 minutes, during which homework outcomes are discussed (15 minutes) and some questions on the theoretical part of the course are asked (30 minutes). The final evaluation keeps into account both the quality of the work done during the course and the performance during the final exam.