Traffic and mobility data collection for real-time applications

Abstract

Successful development of effective real-time traffic management and information systems requires high quality traffic information in real-time. This paper presents the state-of-the-art of traffic and general mobility sensory technology and a suite of methods for data pre-processing and cleaning for real-time applications. We propose a suite of methods and techniques to be applied from traffic data acquisition, preprocessing, transformation and integration until data advanced processing and transfer. Next, we detail some techniques for data preprocessing and integration, or fusion, phases. Even though the comprehensive use of historical traffic data and assignment models to support the most part of online services and operations, real-time data is extremely important to promote models' accuracy and, therefore, the reliability of information and outputs derived from data fusion and processing. Together with techniques and theoretical formulas we present a case study applied to the Portuguese Brisa's A5 motorway, a 25 km inter-urban highway between Lisbon and Cascais. Traffic on this motorway heading to Lisbon in the morning rush hours typically experiences high levels of congestion. Brisa, the motorway operator company, has equipped A5 with a variety of traffic sensors to be used in a real-time multipurpose way, either for traffic management and control or for traveler information and third-part applications.

A Survey on Visual Traffic Simulation: Models, Evaluations, and Applications in Autonomous Driving

“…Traffic sensors come in several forms [LBH*10, Led08]. To list few examples, one fixed sensor is inductive‐loop detector, which is usually placed on highways and major roads to record the attributes of every vehicle that passes.…”

An Overview of Indicators and Indices Used for Urban Mobility Assessment

“…By applying an appropriate methodology, the indices are set-up and used to carry out the mobility assessment. The initial step is to identify and collect data related to the urban mobility [24,25].…”

“…The conventional data sources include the usual sources such as the sensors within the transport infrastructure, meteorological sensors, ecological sensors, etc. [25]. The second group of the data sources encompasses the information and communication systems owned and/or used by the urban mobility shareholders (the organisations providing public transport services, toll collection and similar).…”

“…The objective of the indicator is to represent mobility and trends in a compact, simple and clear way [40]. On the basis of the suitable (available) data from the segments of the transport system, Table 1 -The list of data sources usable for the purpose of the mobility assessment [22,25,26,[28][29][30][31][32][33]38] [48] define the following indicators for the purpose of the mobility assessment: the average transport network speed, the network speed indicator, the average vehicle speed, the vehicle speed indicator, the indicator of time spent within the network, the network delays indicator, the average travelling time, the average density of pollutants generated by transport, the average exposure to the transport-generated pollutants, the average exposure to the overall pollutants, the average duration of the exposure to the pollutants, the total polluting emissions. Norwood et al [6] define mobility by using several key indicators which need to be included within every assessment.…”

Optimal Missing Value Estimation Algorithm for Groundwater Levels

“…Simple (interpolation with last value and mean) and sophisticated methods, such as linear regression and PCA are also applicable in missing data imputation scenario [2]. Tensor-based methods also produce very good results in estimating missing values [3].…”