Methodology
This study sought to quantify the potential benefits of coordinating HST in rural areas because there was a lack of literature with quantified benefits of service coordination. The goal of the research was to define a reasonable expectation of coordination impacts by a few key performance indicators, including: vehicle miles traveled, vehicle hours traveled, passenger load, and passenger trip times. The objective was to simulate and compare potential system performance impacts of program-level trip coordination on transportation providers.

The authors chose two multi-county study areas in North Carolina (KARTS) and South Carolina (SWRTA) and used actual trip request data to simulate different levels of coordination based on the funding sources for the trips. Using the funding sources for each trip, each trip was categorized as Medicaid, Aging, or Other. Medicaid trips were any trips that were funded by the local Medicaid office. Aging trips were any trips funded by the local Councils on Aging and FTA 5310 grants. Other trips were all other requested trips with funding sources that did not fall into the Medicaid or Aging categories.

Two days of actual trip request data were processed and simulated using 3 scenarios: Some Coordination, More Coordination, and Full Coordination. Some Coordination distributed the vehicle fleet proportionally to the three groups based on the number of trips requested for each of the funding categories. The trips within each category were all assigned to vehicles designated for that group and passengers from different funding groups were not commingled. The More Coordination scenario pooled the vehicle fleets and trips for the Aging and Other funding categories and the same process was repeated. Lastly, the Full Coordination scenario utilized the full fleet of vehicles and trips and optimized trips regardless of funding source to represent the ideal coordination scenario.

Findings
The agency in South Carolina (SWRTA) saw 10 to 13 percent improvements in operating efficiency when comparing the Full Coordination scenario to the Some Coordination scenario (Table 1). KARTS, in North Carolina, saw 7 to 9 percent improvements in operating efficiency when comparing the Full Coordination scenario to the Some Coordination scenario (Table 2).

Table 1: SWRTA Aggregate Performance Measures - September 18-19, 2013

Performance Measures	Scenario #1 Some Coordination	Scenario #2 More Coordination	Scenario #3 Full Coordination	% Difference (Scenario 1 and 3)	% Difference (Scenario 2 and 3)
Total Vehicle Hours (hrs.)	1,141	1,105	1,024	-10%	-7%
Total Revenue Hours (hrs.)	1,024	1,000	925	-10%	-7%
Total Vehicle Distance (mi.)	21,634	20,525	18,826	-13%	-8%
Total Revenue Distance (mi.)	17,126	16,519	15,424	-10%	-7%
Passengers per Revenue Hour	1.53	1.57	1.69	11%	8%
# of Vehicles Used	Day 1:  51 vehicles - Day 2: 65 vehicles

Table 2: KARTS Aggregate Performance Measures - September 18-19, 2013

Performance Measures	Scenario #1 Some Coordination	Scenario #2 More Coordination	Scenario #3 Full Coordination	% Difference (Scenario 1 and 3)	% Difference (Scenario 2 and 3)
Total Vehicle Hours (hrs.)	730	700	672	-8%	-4%
Total Revenue Hours (hrs.)	660	630	604	-9%	-4%
Total Vehicle Distance (mi.)	14,989	14,714	13,946	-7%	-5%
Total Revenue Distance (mi.)	12,429	12,173	11,514	-7%	-5%
Passenger per Revenue Hour	1.65	1.73	1.81	9%	4%
# of Vehicles Used	Day 1:  37 vehicles - Day 2: 32 vehicles

Overall, passengers only experienced minimal increases in individual trip times as a result of increased coordination, and in some cases, saw decreases in trip time (Table 3).

	Scenario	Average Trip Time (minutes)
		Some Coordination	More Coordination	Full Coordination
SWRTA	Medicaid Only (45%)	57.8	57.8	54.8
	Aging (19%)	48.6	50.6	53.2*
	Others (36%)	53.9	51.8	52.9
	Total	54.3	53.8	53.6
KARTS	Medicaid Only (33%)	45.2	45.2	51.4*+
	Aging (33%)	47.8	52.2	53.3
	Others (34%)	50.0	53.9	52.9
	Total	47.7	50.5	52.5*

*Statistically significant at 95% level of confidence between Some and Full
⁺ Statistically significant at 95% level of confidence between More and Full