Evaluation of Adaptive Signal Control Technology\u2014Volume 3: Comparison of TBC 2017 and ASCT 2017

2019Volume 3 ReportPDFTech ReportFHWA-ICT-19-003R27-127Adaptive controlAlternatives analysisBefore and after studiesSignalized intersectionsTraffic delaysTraffic queuingTraffic signal control systemsTraffic volumeAdaptive traffic signalsASCTsignalized intersectionscorridor travel timetraffic coordinationfield data on delayqueue lengthHCMsaturation flowsignal timing.IllinoisUnited StatesIllinois Center for TransportationBenekohal, Rahim F.Garshasebi, BehnoushShang, MingfengJeon, HongjaeUniversity of Illinois at Urbana-Champaign. Dept. of Civil and Environmental EngineeringIllinois Center for TransportationIllinois. Department of TransportationUnited States. Federal Highway AdministrationUnited States. Department of TransportationUS Transportation CollectionData were collected at five intersections along the Neil Street corridor in Champaign, Illinois, before deployment of SynchroGreen, an adaptive signal control technology (ASCT). The volume, delay, and queue length data from the field for TBC (time based coordination) 2017 conditions were compared to the data from ASCT (adaptive signal control technology) 2017 conditions, at the 97% confidence level. The field volumes were compared for 57 lane groups (approaches). Traffic volume on 7% of the lane groups significantly increased, on 72% remained unchanged, and on only 21% significantly decreased. Stopped delays increased in 56% of the cases, remained unchanged in 40%, and decreased in 4%. Queue length increased in 35% of the cases and remained unchanged in 65%. To determine ASCT performance, the changes in volume, delay, and queue length combined were considered. An overall performance indicator (PI) was determined for each approach of each intersection at each time period. The performance indicators were Imp (Improved), Unch (Unchanged), and Det (Deteriorated), with 91% confidence. One lane group was excluded from further analysis due to insufficient volume; of the 56 lane groups analyzed, 5% showed improvement, 32% remained unchanged, and 63% (35 cases) showed deterioration. Out of 35 cases, deterioration in 20 cases could be explained by contributing factors such as frequency of unfavorable arrival types under ASCT 2017, as compared to TBC 2017; a few cases of volume increase under ASCT 2017; ASCT miscount of traffic volumes; signal timing changes under ASCT 2017; and increased proportion of vehicles stopped under ASCT 2017. However, in the 15 remaining cases, there was no reasonable explanation for the PI deteriorations when ASCT was operating.80