Research Projects

Stop the Video

Research Projects

STATUS: Complete YEAR: 2017 TOPIC AREA: Public transit, land use, and urban mobility CENTER: METRANS UTC PSR

Smart Sensing System for Real-time Automatic Traffic Analysis of Highway Rest Areas

Project Summary

Project number: PSR-17-11
Funding source: US DOT
Contract number: DTRT13-G-UTC57
Funding amount: $69,197
Performance period: 9/1/2017 to 8/31/2018

Project description

State transportation agency spends millions of dollars annually to maintain and improve the service provided to the drivers in the highway rest areas. In order to collect traffic data in real-time, Researchers can use the vehicle data in the rest areas. Therefore, it is helpful immensely to update the existing safety policies in the rest areas. Transportation agencies don't have any automated systems to perform "automatic" and "real-time" vehicle identification and classification in the highway rest areas. Motivated by a dire need to enhance and modernize the transportation system, we propose an advanced modular system that will integrate a smart sensor to extract a rest area traffic pattern in real-time. Currently, Caltrans collects traffic data from Automated Vehicle Classification (AVC) stations and also manual census collected in the specific locations. However, this technology is too expensive, time-consuming, and disruptive; therefore it has not been used widely in many different locations. 


In recent years, There have been many significant improvements in MEMS sensors domain with respect to size, cost and accuracy. Moreover, extreme miniaturization of RF transceivers and low power micro-controllers have motivated researchers to develop small and low power sensors and radio equipped modules. These sensors are gradually replacing traditional wired sensor systems. These modules which are often called "sensor mote" (size of a quarter) communicate with other sensor nodes and build an intelligent network of sensors. Because of the miniaturization and low power consumption, these sensor motes are extremely efficient due to their low power budget. We propose a wireless MEMS sensor based automatic vehicle classification and identification system for highways rest areas.


Our developed Automatic Vehicle Classification and Identification (AVCI) system consists of two parts, AVCI sensor nodes containing magneto-resistive and accelerometer sensors. These sensors calculate speed and axles respectively. The next part, the system proposes a Access Point (AP) which collects data from sensor motes and calculate speed, axles counts and then it classifies the collected data based on Federal Highway Administration (FHWA) 13-categories Scheme-F[5]. The AP includes a RF transceiver to communicate with the sensor motes and also a GPRS (General Packet Radio Service) shield to transmit aggregated traffic data to the county or regional traffic data collection center.


Mohammad Mozumdar
Assistant Professor, Department of Electrical Engineering; College of Engineering
1250 Bellflower Blvd.
ECS-521Long Beach, CA 90840-8306
United States
[email protected]