06-05-2011, 09:37 AM
[attachment=13320]
1. Project Overview
1.1 INTRODUCTION
The majority of traffic detection technology and sensor research has focused on the
detection of “motorized” traffic. This has left a need for objective information on the
performance of “non-motorized’ traffic detectors. Bicycles and pedestrians are the two
most common types of non-motorized traffic. In recent years, sensors have been
developed to detect the presence of pedestrians and the speed and presence of bicycles.
This has been accomplished with a variety of technologies, including microwave,
infrared, video and inductive loops. With the development of ITS applications,
automated pedestrian detectors are beginning to compliment the existing pushbutton
detectors. These applications optimize intersection operations and improve safety by
reducing the conflicts between vehicles and pedestrians. Similarly, bicycle detectors are
being used to provide detection inputs to traffic signals for call and extend functions.
These technologies are also a valuable tool in gathering historic data. This report
documents the current state of pedestrian and bicycle detection and presents evaluation
results from a field test conducted of six different bicycle/pedestrian sensors.
1.2 BACKGROUND
The Bicycle and Pedestrian Detection Project is a component of the Evaluation of Non-
Intrusive Technologies for Traffic Detection (NIT Project) which is funded by Federal
Highway Administration (FHWA) and Minnesota Department of Transportation
(Mn/DOT). The evaluation of sensor performances for motorized traffic detection was
completed in two phases; Phase I evaluated 17 different non-intrusive sensors from 1995
to 1997 and Phase II investigated 9 non-intrusive sensors from 2000 to 2002. Results
obtained from these two phases of research are available on the Mn/DOT
In October 2001, an Evaluation Test Plan was developed to guide the evaluation of NIT
Phase II. The desire to detect non-motorized traffic was listed as one of the project goals
and objectives in this test plan. After the completion of the Phase II Final Report in
September 2002, the bicycle and pedestrian evaluation started. Literature search, sensor
procurement and test site selection and preparation were conducted prior to the official
data collection and evaluation. Official data collection was conducted on October 8,
2002. Data analysis and report preparation were completed in the subsequent months.
1.3 PROJECT OBJECTIVES
The primary goal of this project is to identify the applications and evaluate the accuracy
of different non-intrusive technologies in detection of non-motorized traffic, namely
bicycles and pedestrians. This project goal is supported by the following objectives:
Objective 1: Identify the applications for non-motorized traffic detection
Objective 2: List similar projects that had been conducted in non-motorized traffic
detection
Objective 3: Conduct a field test to evaluate participating sensor performance
1.4 PARTICIPATING VENDORS AND SENSORS
An extensive vendor database was developed in NIT Phase II. This database includes
vendors from the United States and around the world that manufacture non-intrusive
traffic sensors. The database is contained in Appendices A and B of the NIT Phase II
Final Report (available on the project website). This database was reviewed to select
those sensors that have the ability to detect bicycles or pedestrians. These vendors were
contacted in August 2002 to identify candidate sensors and determine their interest in
participating in this test. Ultimately, five vendors agreed to participate in the test,
including one international vendor. With the addition of existing inductive loop
detectors, a total of six sensors representing five different technologies were field tested
and evaluated.
Table 1 lists the vendors, sensors and other technical data for those that participated in the
field evaluation. Notice that the list includes a combination non-intrusive and intrusive
sensors. Intrusive sensors were included because bicycle and pedestrian trails are much
easier to close than roadways during sensor installation.