Project Number:
04-08
Research Project:
SURE-SE:Sensors for Unexpected Roadway Events:Simulation and Evaluation
P.I. Name & Address:
John Heidemann
Information Sciences Institute
University of Southern California
4676 Admiralty Way
Marina del Rey, CA 90292-6695
Tel:(310) 448-8708
Fax:(310) 823-6714
Email: johnh@isi.edu
Co-P.I.s:
Genevieve Giuliano
School of Policy, Planning, and Development
University of Southern California
Los Angeles, CA 90089-0626
Tel:(213) 740-3956
Fax:(213) 740-0001
Email: giuliano@usc.edu
Project Objective:Objective:
The purpose of this research is to demonstrate the feasibility of an accurate, low cost and rapidly deployable vehicle traffic classification system, a Network Of Traffic Sensors (NOTS).This system will consist of a number of small, low cost computer nodes, each with one or a few sensors (such as pneumatic tubes or adhesive magnetic sensors), connected with a wireless network.Although individual sensors may be relatively inaccurate, we expect this research to develop algorithms that allow the combination of individual sensor readings in the sensor network to provide highly accurate vehicle classification.Our work is complementary with research into new individual sensors, rather than build a new individual sensor, we will show how existing sensors can work together to improve accuracy.Similarly, while we will use current research in radios and energy-conserving sensor networking, the key contribution of our work is developing new algorithms to apply sensor networks to vehicle classification.
Background:
Currently deployed vehicle traffic monitoring systems consistent with either emplaced and relatively accurate sensors such as in-ground induction loops or elevated video cameras, or of deployable but less accurate sensors such as pneumatic tubes.Both have critical limitations: sophisticated, emplaced traffic control systems today can be accurate and are essential to managing traffic flow, but such systems cover only major roadways and cannot be quickly deployed to new areas; they require substantial amounts of investment and planning to extend.Such systems cannot be used for important transient situations, including short-term data collection to model freight flow, traffic management around construction zones, or for emergency situations.Today's deployable systems, on the other hand, are more flexible, but they lack the classification accuracy needed for data collection, and lack the capacity for integration with existing traffic-control systems needed for construction or emergency deployment.
Approach:
The convergence of inexpensive small computers, sensors and wireless networking creates the possibility for sensor networks:collections of many small computers equipped with sensors and radios, providing collaborative and distributed ability to sense, process, and communicate.Sensor networks have several crucial advantages.First, we will develop approaches that allow correlated readings from multiple sensors to reduce measurement error.Second, we will explore algorithms for self-configuration to ensure that a sensor network can be deployed by field crews without manual configuration.Combined with physically movable sensors, self-configuration will allow our network of traffic sensors (NOTS) to be easily moved where it is needed, even on a daily basis.Third, because a NOTS is intelligent, it will be feasible to connect it with local traffic control equipment and centralized traffic management systems.Finally, we will take advantage of current research in inexpensive low-power hardware, and our work in energy-conserving network protocols to ensure that such a system can be battery operated for long periods of time (weeks or more).
The product of our work will be the development of new algorithms (design and computer software)demonstrating correlated readings and self-configuration.We will demonstrate these capabilities through simulations and a system design to prove the feasibility of a NOTS.To parameterize and validate these simulations, we will evaluate sensor accuracy both by surveying existing literature and through laboratory experiments with specific sensors.We expect that research in following years can rapidly translate these algorithms into a prototype system for field tests, then field trials and deployment.
Significance:
The proposed research is significant in using a novel approach to traffic monitoring-rather than improve individual sensors or sensor signal processing, we exploit the computational and communications capabilities of networked sensors to improve accuracy. Our work is complementary with ongoing work in improving individual sensors; we expect to adapt our multi-sensor algorithms to new individual sensors as available.We are aware of no other work exploring the use of multiple sensors.
The primary application of this work will be the development of a system for collection truck traffic data to support research in commercial goods movement (Topic Area 1).Additional applications include deploying NOTS around a construction area to improve worker safety (Topic Area 4), NOTS can be deployed to assist traffic flow in emergency situations such as major evacuations (Topic Area 4)
Task Descriptions:
1. Review current traffic sensor systems.
2. Purchase and evaluate selected sensors.
3. Design multi-sensor classification algorithms.
4. Evaluate multi-sensor classification algorithms through simulation.
5. Revise algorithms based on simulations.
6. Laboratory prototype of system (if possible) based on simulation results.
7. Evaluate deployability of NOTS.
8. Write final report.
Milestones, Dates:
January5, 2004 - January 4, 2005
Total Budget:
$75,000
Student Involvement:
One Student @ $21.51/hr, 450 hours
Relationship to Other Research Projects:
New topic area, part of goods movement focus area
Technology Transfer:
Project report will be posted on the website; cooperative effort with Los Angeles Department of Transportation
Potential Benefits of the Project:
Better information on truck flows
TRB Keywords:
Vehicle classification; sensor networks
Primary Subject:
4b.9 Traffic management
Goals:
Mobility
Enabling Research:
Sensing and measurement
Modal Orientation:
