In response to Onondaga County not meeting air quality standards in 1993, the City of Syracuse implemented the Signal Interconnect Design Project. The project standardized 145 intersections within the City of Syracuse and optimized the signal timing in an attempt to reduce automobile emissions by creating a more efficient network.

The study area included 37 of the 145 intersections, all located along five main arterials. The analysis involved comparison of simulation results representing conditions before and after implementation of the coordinated signal timing. The Synchro™ software package was used to model the performance of the system before and after the improvement project. The before data were collected in 1998 and consist of traffic counts for three time periods (AM, Mid-day, PM), as well as signal timing data. Field travel time data was collected, both before and after implementation, to confirm that the times estimated by the simulations accurately represented actual travel times.

The network simulation was run multiple times to optimize the signal timing. Once the optimized configuration was determined, the signal timings were implemented into the field controllers. To account for real-world conditions not accurately represented in Synchro™, staff calculated field adjustments to the traffic signal timing plans by driving the arterials multiple times at all time periods. These final adjustments, implemented in early 2000, were then input into the Synchro™ package, the simulation tool was run for the thirty-seven intersections, and the performance measures were calculated.

Traffic counts used in the simulation model for both the before and after case were collected in 1998. The authors felt that the 1998 data accurately represented after conditions due to little to no development or population changes in the study area.

Results of the simulations representing before (1998) and after (2000) conditions are summarized below.

Measure of Effectiveness	Unit of Measure	AM PEAK PERIOD		MID DAY PEAK PERIOD		PM PEAK PERIOD
		Before	After	Before	After	Before	After
Signal Delay	seconds/veh	14	12	13	11	16	13
Total Stops	each	26,411	23,467	19,840	16,734	30,003	25,236
Average Vehicle Speed	miles/hour	14	15	13	15	12	14
Total Fuel Consumption	gallons	490	447	367	328	546	475
Average Fuel Economy	miles/gallon	10.4	11.1	10.3	11.4	9.4	10.7
Average Vehicle Emissions	Kg CO	34.24	31.27	25.63	22.95	38.15	33.2
	Kg NOx	6.66	6.08	4.99	4.46	7.42	6.46
	Kg VOC	7.94	7.25	5.94	5.32	8.84	7.69

Measure of Effectiveness	Unit of Measure	Range of Improvement (by intersection)		Range of Improvement (entire network)
Total Vehicular Delay	seconds/veh	7.14%	80.95%	14.29%	18.75%
Total Stops	each	-3.84%	84.81%	11.15%	15.89%
Average Vehicle Speed	miles/hour	0.00%	66.67%	7.14%	16.67%
Total Fuel Consumption	gallons	2.56%	40.00%	8.78%	13.00%
Average Fuel Consumption	miles/gallon	2.46%	63.49%	6.73%	13.83%
Average Vehicle Emissions	Kg CO	2.58%	41.35%	8.67%	12.98%
	Kg NOx	3.70%	40.00%	8.71%	12.94%
	Kg VOC	3.17%	41.67%	8.82%	13.01%

The study observed improvements in most system performance measures for the entire network over all three time periods. When analyzing each intersection individually, a large range of improvement is observed and, in some cases, there is no improvement or decline. The upper and lower bounds on the ranges of improvement measured by intersection were varied by time period for all metrics. Notably, when observing the ranges of improvement for the entire network, the lower bound for all metrics is found exclusively in the AM time period and the upper bound for all metrics is found exclusively in the PM time period.

Improvements in travel time on each of the five main arterials ranged from 1.15 to 34.20 percent during the AM peak, -2.71 to 35.10 percent during the Mid Day period, and -13.88 to 31.18 percent during the PM peak. Level of Service (calculated by seconds of delay) was calculated for 41 intersections, at three different times of the day, for a total of 123 measures. Thirty-six of these measurements improved, 73 remained the same and 14 declined.