Freight transportation is central to economic growth and much of this advancement is pegged to commercial vehicle transportation. However, with the ever-rising concerns of traffic congestion, air pollution, safety, and climate change, it is critical to understand commercial vehicle activity in a bid to mitigate its effects. Transportation agencies require accurate and detailed commercial vehicle data in order to predict freight movement and devise policies and investment strategies that will minimize the adverse impacts freight transportation in the areas mentioned. However, in the US such pivotal data are either missing or costly to obtain from the current data sources. The current sources do not provide detailed truck activity information such as commodity carried, vehicle configuration, and industry affiliation.
Dr. Sarah Hernandez from the University of Arkansas in collaboration with Dr. Andre Tok, and Professor Stephen Ritchie at University of California Irvine developed an approach for obtaining data from two traffic sensor technologies – Inductive Loop Sensors and Weigh-in-Motion Sensors (WIM) – by leveraging existing WIM detector infrastructure and subsequently integrated the data to develop an advanced truck body classification model using a multiple classifier framework. Because of the strong complementary characteristics of these two sensor technologies – WIM data provides accurate axle measurements while inductive signature data captures detailed profiles that can be used to discern truck chassis types – the system can help to collect high resolution truck data relating to axle and body configuration for better understanding of industry and freight trucking activity. Their work is published in the journal, Transportation Research Part C.
The researchers first developed an interface to integrate inductive signature detectors with an existing weigh-in-motion controller. Signature data was logged into an on-site field computer through a USB port at the front face of the signature detector card, while axle spacing and weight data was logged via a separate serial interface cable between the WIM controller and the field computer.
The integrated system allowed the authors to collect both the weight-in-motion and inductive signature data concurrently. For every vehicle traversing through this integrated system, the authors could concurrently collect inductive signatures from the inductive loop detectors while axle configuration and weight were obtained from the bending plate sensors. The data streams were then post processed to provide input data to the truck body model.
Using the integrated data sources, the authors developed an ensemble-based advanced classification model to distinguish over 60 truck configurations that can relate to the types of commodity hauled and industry served. The approach provided in this paper can enable transportation agencies to effectively improve on air quality control, freight planning, and management of infrastructure. For freight planning, specifically, the integration of weight measurements with predicted body configuration information provides a means to estimate truck payloads and empty factors that serve as indicators of freight efficiency and provide critical inputs for long-range commodity-based freight forecasting models.
“This paper represents the confluence of two prevalent transportation research areas: freight transportation planning and intelligent transportation systems. We uniquely use existing ITS technologies to fill critical data gaps identified in long-range freight forecasting and air quality models.” Said Andre Tok
Sarah V. Hernandez, Andre Tok, and Stephen G. Ritchie. Integration of Weigh-in-Motion (WIM) and inductive signature data for truck body classification. Transportation Research Part C, volume 68 (2016), pages 1–21.
Go To Transportation Research Part C