Judson S Matthias

EVALUATION OF AN ALTERNATIVE TRAFFIC LIGHT CHANGE ANTICIPATION SYSTEM

A driving simulator was used to study the efficacy of an alternative signal phasing program. The new Traffic Light Change Anticipation System (TLCAS) utilizes a flashing amber in conjunction with a solid green indication to warn drivers of the impending onset of the solid amber indication. This new program was expected to provide drivers with additional information with which to make safe stopping or crossing decisions, and reduce behaviour associated with increased accident rates at signalized intersections. Additionally, a new measure of first response variability was introduced to evaluate the effect of the new program on driver behavior. The results indicated that the new system has the ability to reduce the number of red light violations. Of the 1148 target intersections, 43 violations were recorded for the regular program compared to 9 for the TLCAS program. The severity of decelerations was also impacted. The regular program displayed average maximum decelerations of 3.1 m/sec2 compared to 2.5 m/sec2 for the TLCAS program. However, the TLCAS program showed an increased variability in first response five times larger than the regular program. This finding, in conjunction with traditional measures, indicates that the new system performs comparably to an increased amber duration by increasing the potential for conflicting decisions between successive drivers approaching an intersection. Altogether, the results suggest that this alternative signal phasing program would not improve intersection safety. The findings also suggest that further study of the longitudinal behavior after the introduction of a TLCAS program may be warranted. Specifically, future studies should examine whether subjects engage in appropriate adaptive changes when confronted with a solid amber period warning after becoming familiar with the TLCAS.

Simulator Evaluation of Green and Flashing Amber Signal Phasing

A research study conducted to evaluate the efficacy of flashing amber signal phasing is reported. Flashing amber, which is set to overlap with the green indication a few seconds before the onset of solid amber, is a form of time reference aid used to warn drivers of an impending onset of amber. The time reference aid is a concept predicated on the principle that driver decisions will be easier and more predictable if drivers have advance information that allows them to predict the onset of amber. The evaluation of the flashing amber method showed that its implementation has the potential to reduce red-light violations, severity of maximum decelerations, and kinematically defined inappropriate stop or cross decisions. However, the data also showed that compared with the regular signal phasing, the flashing amber phasing increased the size of the indecision zone, a mechanism usually responsible for increased rear-end collisions. A measure not previously used in literature was introduced to compare the regular and the experimental signal phasing. This measure, which analyzes first-response time variability in relation to the indecision zone, predicted that increased rear-end collisions might be expected as a result of implementation of the flashing amber signal phasing. Generally, the results suggest that the implementation of a flashing amber signal phasing would not significantly increase intersection safety despite the notion that it would improve driver anticipation of the onset of solid amber.

MODELING DETECTION OF INCIDENTS BY DRIVERS

Abstract An analytical model was developed to assess the detection of freeway incidents by drivers using advanced communications technologies. Drivers can report incidents to a highway agency by voice or by digital messaging either directly or through roadside beacons using a fixed or portable wireless communications system such as a cellular telephone. A FRESIM model was used to simulate shoulder and lane-blocking incidents occurring under variable traffic flow rates. The proportion of drivers with an in-vehicle communication medium arriving upstream of an incident and the degree of the drivers’ reporting propensity were introduced as controlled variables. A discrete probability model was used to represent binary reporting decisions of drivers arriving upstream of an incident. The results showed that all incident types that were simulated could be detected in a short time with high probability of detection. The data indicated that all incident types, regardless of the prevailing traffic volume at the time of their occurrence, were detectable in less than one minute if one of five drivers on a freeway had an in-vehicle communication medium and was willing to use it to report the incident. Of special importance is that at the end of 1995 there was one cellular telephone user per five licensed drivers in the United States. The results of the modeling further showed that other factors affecting the detection performance of a driver-initiated incident detection process were the willingness of drivers to report incidents, the length of roadway within which an incident is visible to a driver, and the austerity of the highway agency’s procedure for mitigating false reporting.

A traffic assignment planning model: The load-node concept

Abstract The use of the traffic assignment model in urban transportation planning is universal; however, traffic assignment in its present form is expensive, cumbersome, and has serious limitations. In this research the authors have developed and tested the load-node concept of traffic assignment which reduces many of these problems. In the load-node concept the corners of zone boundaries are used for points of trip origins and destinations. In this manner the zone centroids and associated connector linkages of the network are removed. This results in a substantial reduction in network coding. In this study the total number of nodes required to define the network decreased by almost 3000 to 995 and the number of links in the network decreased by two-thirds. The forecasted 1990 Phoenix metropolitan area was used to test this approach and the average link volumes of all-or-nothing capacity restraint assignments using both the load-node and zone centroid concepts formed the basis of the comparison. The analyses indicated the average link volumes of the two concepts are comparable. A probable reduction in computer time of 40 + per cent was obtained when the load-node concept was used. The authors suggest that this approach is primarily adaptable for planning purposes and not design. With respect to the design use of traffic assignment results, the travel between adjacent zones is seldom included in the major arterial street and freeway traffic volumes obtained from zone centroid assignment. This may result in a substantial underestimation of future traffic volume and turning movements particularly if stepped down 24-hr assignments are used for design purposes. The load-node approach does assign travel between adjacent zones.

Validation of Left-Turn Delay at Two-Way Stop-Controlled Intersections

Control delay for left-turning vehicles at unsignalized intersections was observed in the field and compared with average control delay calculated from the methodologies presented in the 1997 update of the Highway Capacity Manual (HCM). Unsignalized intersections with two-way left-turn lanes on the major street were observed in the peak and offpeak hours, and control delays were recorded for the one-stage and twostage left-turn processes. Next, the methodologies presented in the HCM were used to calculate the control delay for both processes and compared with the observed data. These comparisons were used as the basis for validation of the HCM methodologies regarding left-turn control delay at unsignalized intersections. From the comparisons, the calculated delay closely corresponds with the observed data, with a total approach volume at the intersection of approximately 2,500 vehicles per hour or less. Once the total approach volume increases above this level, the calculated values rapidly increase and the actual observed control delays gradually increase at a much lower rate. As a result, the observed and calculated delays are different when the intersection handles more than 2,500 approach vehicles in an hour. Statistical analyses were performed on the data to determine if the average observed control delay was related to the calculated control delay. Statistically, the observed control delay and the calculated control delay at the 95 percent confidence level show that the two data sets yield similar results for off-peak conditions. However, during the peak hour, when the total approach volumes are higher, the 95 percent confidence interval yields different results. Hence, the HCM procedures produce, on average, greater control delay estimates than the field observations when the total approach volumes are high.