A study of the left-turn-on red (LTOR) traffic light scheme in Singapore
Date of Issue2015
School of Civil and Environmental Engineering
Since 1997, the LTA has been implementing the Left-Turn-On-Red (LTOR) traffic control scheme at selected signalised junctions in an attempt to increase left-turning capacity and reduce stationary time delays. The number of implementations numbered 44 in 2003. Currently, only 16 remain mainly because of a change in junction design and a general increase in pedestrian traffic in Singapore. In the past, this scheme has generally been shown to improve flow and lessen waiting time. However, a vehicle turning left on a red light theoretically poses an additional risk to some parties. Theoretically the main at-risk parties are a) pedestrian on a crossing with the green man showing, directly in front of the left turning vehicle and; b) the oncoming traffic heading straight on the adjoining roadway. The objective of this project is to assess both the effectiveness of the scheme in reducing delays at signalised junctions and the propensity of LTOR to encourage the above stated theoretical associated risks. Two signalised junctions were selected for this study. A field survey was done at each site by gathering footage and reviewing it on a monitor. Data gathered included information on general vehicle and pedestrian numbers, vehicle movement and signal compliance and road user interaction as a result of LTOR. It was discovered that LTOR is effective at providing the opportunity to turn for more than half left-turning vehicles before the green light as long as a dedicated lane is provided. At the junction using a shared left turn lane where only 1/3 of the traffic was turning left, little benefits were seen. It was observed that <10% of motorists were still not aware of the existence of this scheme as they failed to turn given appropriate circumstances. A key issue with the safety of this scheme is the fact that over 80% of vehicles which manage to turn on red violate the scheme by stopping beyond the stop line. 1/3 did not stop at all. Another problem noted includes the highly prevalent tendency of vehicles to creep into the junction unnecessarily. The number of cycles containing both non–compliant drivers and pedestrians was recorded. In total this overlap occurred in 16% of cycles. Over the 3 hours, that resulted in a maximum possible number of 25 opportunities for a collision to occur. There were 3 conflicts where the LTOR vehicle had to stop harshly and 1 conflict where a pedestrian chose to run across the junction. In all, the LTOR scheme clearly provides a benefit to left-turning vehicles. However, given that the scheme is necessarily restricted to those few junctions with low vehicle and pedestrian volume, its utility is questionable. The accident risk especially in these locations is undoubtedly as marginal as the advantages of the scheme.
Final Year Project (FYP)
Nanyang Technological University