Comprehensive Freight Demand Data Collection Framework for Large Urban Areas - Sustainable Urban Logistics: Concepts, Methods and Information Systems - page 92

Equally important when describing the system is to characterize the transpor-
tation modes, as they determine the transportation technology and infrastructure
used, and operational needs. A clear understanding of how the goods are trans-
ported would help during the design of a data collection procedure when identi-
fying who, where and how to survey.
As shown, characterizing the freight system could be a challenging endeavor
due to the many facets that should be considered when designing a freight demand
data collection framework. Building on the discussion about the components of the
freight system, the following section analyzes the type of data required for freight
transportation modeling and discusses the possible data sources.
3 Identification of Data Needs and Sources
Designing a comprehensive data collection framework requires the use of a sys-
tematic approach. In doing this, the initial step is to identify the data requirements
for the freight modeling technique that would be used. To simplify the exposition,
the authors used a classification system based on: modeling focus, modeling
principle, and flow unit. In this context:
• Modeling focus, which could be trip interchanges linking an origin to a desti-
nation; or tours, i.e., a sequence of delivery stops.
• Flow unit, which could be commodity flow, value, vehicle trips, or other.
• Modeling principle, which refers to the techniques used in modeling.
Table 2 shows the model classification produced. It is important to mention that
the table is by no means comprehensive, nor it is intended to be, as the ones listed
only represent the main types. The endless combinations of modeling possibilities
simply cannot be considered for reasons of space. When analyzing the process of
model development and assessing data requirements, care must be taken to
properly consider the fundamental structure and empirical foundation of the
model, the computational algorithms and data structures, and the process to ana-
lyze modeling results. In general, the data or information required to address these
issues fall under two main categories: data for model calibration (C) and data to
make forecasts (F). For the sake of brevity and conciseness, the authors decided to
identify the key data categories that are typically needed to develop, calibrate, and
do forecasting of freight demand. Brief descriptions of these categories are pro-
vided next.
Information/insight into logistical patterns of flows. Developing a good freight
demand model requires a basic understanding of the functioning of the system
being modeled (i.e., agents, interactions, functions).
Freight generation data. Refers to the study of production and consumption of
freight. It focuses on the analyses and quantification of the transactions between a
producer of cargo and the next consumer.