IMPROVED TECHNIQUES FOR COLLECTING SPEED DATA -- FINAL REPORT
A field evaluation of speed data collection techniques produced the following results: (1) Six vehicle selection strategies were tested in order to eliminate observer bias. All yielded statistically equivalent results to real traffic speeds; however the Randomized (Designated Vehicle; Non-lane specific) strategy consistently produced mean and 85th percentile speed error of .2 mph or less. (2) The accuracy of manually timed speed measurement was determined via vehicle-by-vehicle comparisons aganist automated equipment. Individual vehicle measurement error was generally compensating, and averaged between-observer results produced mean and 85th percentile speed within .5 mph accuracy. (3) The reliability of two methods (radar and manual timing) was assessed at four highway types. Radar produced acceptable accuracies; however manual timing was shown to be reliable only in highway settings which afford an elevated vantage point. (4) Spot speed measurement precision was examined for a variety of effects (i.e. minimum sample, observation duration, observer fatigue). Two minimum requirements were determined: 45-minute observation and 100-vehicle sample. Under these conditions, all-vehicle speed estimations were accurate within 1.0 mph and no manual-timing accuracy degradation resulted due to observer fatigue. (5) Driver slowing response was determined for selected speed sensing devices. No effect was shown for pavement markings, piezoelectric cables, and road tubes. Slowing did occur with radar useage. (6) The effects of speed variation over time and within homogeneous highway sections were studied. Mid-week speeds were found to be slowest; mean speed variation ranging from 4 to 6 mph did occur within highway sections. (7) Determination of a free-flow headway criterion based on platoon stability revealed that headways larger than 4.0 seconds connote free-flowing vehicles.
Transportation Research Corporation2710 Ridge Road
Haymarket, VA United States 22069
Washington, DC United States 20590
- Hanscom, F R
- Publication Date: 1984-11
- Pagination: n.p.
- TRT Terms: Accuracy; Data collection; Headways; Methodology; Radar; Sensors; Traffic speed
- Subject Areas: Highways; Operations and Traffic Management; Safety and Human Factors; I81: Accident Statistics;
- Accession Number: 00457626
- Record Type: Publication
- Report/Paper Numbers: DTFH61-82-C-00064
- Files: TRIS, USDOT
- Created Date: Aug 31 1986 12:00AM