Turner-Fairbank Highway Research Center.

Safety Research > Evaluations of Low Cost Safety Improvements Pooled Fund Study > A Human Centered Systems Research Perspective for Road Safety

PPT version for Printing

A Human Centered Systems
Research Perspective for
Road Safety

Dr. Thomas M. Granda
United States Department of Transportation
Federal Highway Administration
Office of Safety Research and Development
Human Centered Systems

Low-Cost Safety Improvement
Technical Advisory Committee Meeting
June 19, 2007

Overview

Introduction
Description of Human Centered Systems laboratory facilities
Description of nature and scope of research
Selected areas/studies of behavioral research

Picture 1

Picture 2

Picture 3

Picture 4

Human Error & Crash Causation

Turner-Fairbank Highway Research Center HCS Research Capabilities

Picture 1 Sign Simulator				Picture 2 Desktop Simulator
Picture 3 Field Research Vehicle				Picture 4 Highway Driving Simulator(HDS)
Picture 5 Field-Tests,Validation				Picture 6 Photometric Visibility Laboratory

Sign Simulator

Determine meaning of signage
Obtain initial estimates of recognition distance

Low Fidelity Simulation	Field Data Collection
Picture 1	Picture 2
Picture 3	Picture 4

Field Research Vehicle

Picture 1

Picture 2

Picture 3

1999 Saturn SL-1
Data collection system
- 4 cameras
- Video quad unit
- 3 axis digital accelerometer
- GPS receiver
- Inverter for 120v AC power
- Laptop computer for data storage
- Analog and digital video recording

Highway Driving Simulator

Saturn Car cab with 3 df motion base
270 degree field of view
Rear and side view mirrors
cluster-based PC workstation environment
GeForce NVidea 7900 GTX Pro graphics cards
In-house programming using open source code in a Linux Environment

Proper Fit of Research Tools & Activities


Generic Transportation HF Research Activities	SQ	TA	CM	PT	LF	HF	FT
Initial user concept acceptance	•
Initial user preferences	•
Initial user requirements	•	•
Initial evaluation of device parameters		•	•	•
Evaluation of rapid prototype(s)			•	•
Initial evaluation of a procedure			•	•	•
Refine user requirements in tasking situation				•	•
Explore elementary non-driving component				•	•
Non-driving subsystem and driving integration					•	•
Non-driving subsystem and component and driving integration						•
Initial integration of multiple subsystems						•
Study of complex multiple independent variables						•	•
Integration of multiple systems						•	•
Full prototype testing and evaluation						•	•

Research tools proceeds from simple on the left to complex on the right
SQ = Survey/Questionnaire, TA = Task Analysis, CM = Computer Model, PT = Part Task Simulator, LF = Low Fidelity Simulator, HF = High Fidelity Simulator, FT = Field Test

HCS Program Areas

Intersections
- Warning to potential crash victims
- Diverging Diamond Interchange
Pedestrians & Bicycles
Speed Management
Visibility
Operations
- Transportation Management Pooled Fund Study
- Traffic Control Device Pooled Fund Study

Picture 1

Picture 2

Picture 3

Intersections: Infrastructure Based Red-Light Violator Warning

Test 3 warnings in HDS
Assess driver response to unexpected warning
64 % of drivers responded in a way that would prevent collision with a red-light violator

Assessed unexpected warning on Smart Road
90% of drivers stop given 2.7 sec. advance warning
Need to assess driver response when subject vehicle is part of a platoon

Intersections:
Red-Light Violator Warning to Drivers in Platoons

Four platoon scenarios

Lead vehicle brakes
Lead vehicle does not react to warning
Vehicle following participant
No other vehicles present

Three warning distances (assuming 45 mph)

180 ft (2.7 s,)
142.5 ft (2.2 s)
105 ft (1.6 s)

Pedestrian Exposure to Risk

Problem – Crash statistics must be adjusted for exposure – Risk = Crashes / Exposure
Procedure – Conducted literature review of 100 papers – Analyzed previously suggested exposure metrics – Recommended new metric: hundred million pedestrian feet of roadway travelled – Tested feasibility of collecting data with new metric at 7 sites in Washington, DC

Pedestrian Exposure to Risk: Sampling Sites

Calm stop-controlled intersection
Metro stop mid-block crossing
Dead-end street with car repairs, etc.
Busy entertainment area intersection
Busy Metro stop intersection
Shopping mall parking lot
Residential street with many driveways

Pedestrian Exposure to Risk: Findings

It is feasible to collect data using proposed metric
8 hour samples at 7 sites in one city yielded total of 1.13 million pedestrian feet of roadway traveled
Developed special measurement techniques for driveways and parking lots
Proposed metric has promise to serve as denominator of pedestrian risk equation

Pedestrian Exposure to Risk:Next Research Steps

Collect more samples in a single city
Test collecting samples in a rural area
Generalize to entire city or area
Employ combination of empirical sampling and statistical modeling
Generalize to exposure for entire nation
Develop validation and update procedures

Pavement Markings for Speed Reduction: Research Goals

Determine a low-cost pavement marking treatment with a high probability of success
Determine the effectiveness of the marking through a field evaluation at three different locations

Picture 1

Picture 2

Picture 3

Pavement Markings for Speed Reduction: Results

Peripheral Transverse Lines as a Pavement Marking Pattern have the ability to significantly reduce vehicle speeds (4 mph reduction in average speeds in New York and 5 mph reduction in Mississippi)
Appear to be most effective with unfamiliar drivers and in situations where roadway geometry requires a reduction in speed