Hancock Area Transportation Services Optimization
Location
Ada, Ohio
Start Date
9-12-2025 2:00 PM
End Date
9-12-2025 2:10 PM
Description
Hancock Area Transportation Services, otherwise known as HATS, is the public transportation system for Hancock county. It currently operates as a contracted or reservation transportation service, where people must request a ride in advance. However, demand for public transportation in Hancock county has increased. From 2022 to 2025, the number of trips made by HATS per year has increased from 63,890 to approximately 87,463. Additionally, the number of unique passengers and wheelchair rides have doubled to approximately 3,200 per year and 11,050 per year respectively. However, they have been unable to keep up with this increase in demand, this year, an estimated 4,255 rides will have been denied by HATS due to insufficient capacity.
HATS has proposed various potential solutions to this problem. The first is to move to standard fixed-route bus service within the city of Findlay, creating a number of bus routes between many of the busiest destinations and housing areas. Another solution proposed was to create a “microtransit” service with on-demand service within the city of Findlay, maintaining a fleet of vehicles that would operate within specific areas of Findlay and would allow people to request service when needed instead of having to schedule in advance. Finally, they proposed simply improving the existing service with better scheduling, which could also allow people to request transportation with less notice, however, this solution would still require some amount of advance notice.
Although the first solution, fixed bus routes, would likely provide the best service, it would require a significant upfront investment in infrastructure, and it may be difficult to find funding for the service. Based on the reports made by HATS and the data they collected in a survey of HATS users, the current system is plagued by difficulties in scheduling and creating routes for their fleet to successfully complete the user’s trips in a timely manner. Additionally, even if they chose to move forward with the “microtransit” or improved on-demand service, they will need better scheduling capabilities.
This project attempts to utilize machine learning to optimize routes and schedules for a scheduled or on-demand bus transportation service. The goal is to reduce the number of trips that have to be rejected and improve travel times for users. This project achieves this by determining the travel time between all potential stops for the day being scheduled, then attempting to find routes which will successfully transport the users from their origin to their destination while minimizing the travel times for each user and the number of routes needed.
Unfortunately, transportation is a complicated problem where numerous factors must be considered simultaneously. As a result, there has been difficulty attempting to define the problem in a manner in which a machine learning algorithm can attempt to calculate a solution. Currently the program is able to find the travel times between any given street addresses, however, it should be possible to use this data to calculate optimal routes.
Recommended Citation
Freer, Samuel, "Hancock Area Transportation Services Optimization" (2025). College of Engineering Student Research Colloquium. 12.
https://digitalcommons.onu.edu/eng_student_research_colloquium/2025/Presentations/12
Hancock Area Transportation Services Optimization
Ada, Ohio
Hancock Area Transportation Services, otherwise known as HATS, is the public transportation system for Hancock county. It currently operates as a contracted or reservation transportation service, where people must request a ride in advance. However, demand for public transportation in Hancock county has increased. From 2022 to 2025, the number of trips made by HATS per year has increased from 63,890 to approximately 87,463. Additionally, the number of unique passengers and wheelchair rides have doubled to approximately 3,200 per year and 11,050 per year respectively. However, they have been unable to keep up with this increase in demand, this year, an estimated 4,255 rides will have been denied by HATS due to insufficient capacity.
HATS has proposed various potential solutions to this problem. The first is to move to standard fixed-route bus service within the city of Findlay, creating a number of bus routes between many of the busiest destinations and housing areas. Another solution proposed was to create a “microtransit” service with on-demand service within the city of Findlay, maintaining a fleet of vehicles that would operate within specific areas of Findlay and would allow people to request service when needed instead of having to schedule in advance. Finally, they proposed simply improving the existing service with better scheduling, which could also allow people to request transportation with less notice, however, this solution would still require some amount of advance notice.
Although the first solution, fixed bus routes, would likely provide the best service, it would require a significant upfront investment in infrastructure, and it may be difficult to find funding for the service. Based on the reports made by HATS and the data they collected in a survey of HATS users, the current system is plagued by difficulties in scheduling and creating routes for their fleet to successfully complete the user’s trips in a timely manner. Additionally, even if they chose to move forward with the “microtransit” or improved on-demand service, they will need better scheduling capabilities.
This project attempts to utilize machine learning to optimize routes and schedules for a scheduled or on-demand bus transportation service. The goal is to reduce the number of trips that have to be rejected and improve travel times for users. This project achieves this by determining the travel time between all potential stops for the day being scheduled, then attempting to find routes which will successfully transport the users from their origin to their destination while minimizing the travel times for each user and the number of routes needed.
Unfortunately, transportation is a complicated problem where numerous factors must be considered simultaneously. As a result, there has been difficulty attempting to define the problem in a manner in which a machine learning algorithm can attempt to calculate a solution. Currently the program is able to find the travel times between any given street addresses, however, it should be possible to use this data to calculate optimal routes.