Dan R. Middleton
Published: 1994
Total Pages: 168
Get eBook
Providing effective roadside warning devices for drivers of large trucks is critical on freeway connectors where speeds are relatively high but design speeds may be substantially less than on mainlanes. Identifying and testing appropriate methods of monitoring traffic on freeway connectors was also included in this research. Two monitoring systems evolved, one using roadway sensors and the other using roadside sensors. Roadway sensors consisted of both piezoelectric and inductive loop sensors, while roadside sensors applied infrared sensor technology. The roadway warning devices tested can be categorized as passive devices and active devices. Passive devices consisted of "truck tipping" warning signs, while the active device consisted of flashing lights mounted one above and one below a set of passive truck tipping signs on both sides of the roadway. Speed reduction, as associated with accident reduction, was the ultimate goal of these tests. The null hypothesis tested by ANOVA of no treatment effect in the presence of initial speed was rejected in all but one of four models, using the probability of a Type I error, $\\\\alpha, $ equal 0.05. Speed reductions due to the active system were significant in AC and BC data sets, but not in AB data sets, suggesting that truck driver response to the lights occurred downstream of Location B. Truck weights were not significant in any tenable test results, and separation of trucks into the categories of combination and non-combination trucks, peak/off-peak periods, and day/night/dusk periods was not helpful in understanding variations in truck speeds. Cumulative speed distributions showed that the fastest trucks decreased their speeds by approximately 2 to 3 mph (3 to 5 kph) during the test period. Five of the seven single-vehicle truck accidents that were recorded on the I-610/US-59 connector in an 8 1/2 year period that were speed-related resulted in rollover. None occurred after installation of warning treatments being tested, although there were other prior years with no recorded accidents