Poster
Client Context
Peachtree Presbyterian Church (PPC) is the largest Presbyterian church in the United States that welcomes an average of 2,100 visitors on a Sunday. Chartered in 1919, PPC moved to its current location on Roswell Road in 1960 about six miles north of Georgia Tech.
To accommodate the influx of visitors each Sunday morning, the church owns 940 parking spaces between three main lots on the church’s campus: the Dogwood Parking Deck, the Rear Surface Lot, and the Magnolia Lot. The Dogwood Parking Deck and the Magnolia Lot are both located across Roswell Road from the sanctuary, whereas the Rear Surface Lot is adjacent to the sanctuary. In addition to the lots on campus, PPC offers a shuttle service to off-campus sites at Cates Center and Powers Ferry Square. PPC conducts three services at 8:45am, 10:00am, and 11:15am that each last approximately one hour. Additionally, PPC conducts concurrent community events in a separate building during each of the three services. The heaviest congestion occurs between the two most populous services: 10:00am and 11:15am. Cars that cannot find a parking spot in a timely manner often arrive late into the sanctuary, become frustrated, or sometimes decide to leave. Furthermore, circling cars cause traffic to back up, creating gridlocks that extend onto Roswell Road.
Project Objective
To establish that a clear problem exists, PPC administered a parking survey on February 16th to congregants at each service. Approximately 600 surveys were collected on February 16th from congregants attending the three services. More congregants are not satisfied as services continue throughout the day, with 17% dissatisfied at 8:45am and 31% dissatisfied at 11:15am. It is important to rectify this dissatisfaction because unhappy guests may become frustrated, leave the campus to return home, or decide not to return to PPC.
The main goal of this project is to reduce congestion on the church’s campus and Roswell Road. This reduction will minimize the visitor’s transportation time to the sanctuary and eliminate preventable traffic. In turn, the enhanced parking experience will encourage visitors to return to PPC each Sunday. Additionally, reducing the idle time of shuttle buses is the main goal for improving the shuttle bus system.
One of the main points of congestion originates from congregants at the 8:45am and 10:00am services. Those guests who generally do not leave their parking space before the start of the next service reduce the amount of available parking spaces for those incoming for the next service. Another cause of congestion is the uneven utilization across the parking lots. While the Rear Surface Lot and Gym Lot fill and cause traffic, the Magnolia Lot and Dogwood Parking Deck remain with many empty spaces. Instead of finding empty lots, cars circle around the full lots looking for an empty parking spot. This circling wastes time for visitors, and traffic increases as more people seek spaces in a full lot. Therefore, the church continues to witness preventable traffic each Sunday morning.
Next, PPC has two designated parking lots on the opposite side of Roswell Road from the sanctuary. Currently, visitors have three crosswalk options, all with police officers to pause and restart traffic flow when congregants enter and leave the crosswalks. Only one of these crosswalks is painted and coincides with the main traffic light. The other two are convenient places north and south of the main traffic light where Atlanta City Police allow visitors to cross Roswell Road.
Finally, the bus system currently lacks a consistent schedule. While bus drivers can leave each shuttle lot at his or her discretion, congregants have little insight as to when buses will arrive or depart at each location. Moreover, buses will often sit idle with their engines on for over 15 minutes at a bus stop between services.
Design Strategy
The following steps summarize the team’s approach:
1) Research how to replicate the PPC system with a simulation model to address the observed opportunities.
2) Collect data for corresponding parameters in the simulation model.
3) Review, clean, and analyze the collected data. Determine distributions for simulation parameters.
4) Input data parameters into the simulation model and validate the model.
5) Organize a list of possible solutions to alleviate traffic at PPC. Establish a clear success metric. Create a plan for how to simulate each solution.
6) Run each proposed solution in the simulation model. Compare results and rank solutions according to the success metric.
7) Present results and recommendations to the client while noting possible disadvantages of each solution. Share deliverables with the client.
The team decided that a simulation model would be the best way to replicate all of the events that occur on PPC’s campus. To create the most accurate simulation possible, the team collected data in the most significant locations of the PPC system across two months.
The simulation was verified by observing cars within the simulation making decisions similar to that of people driving in real life. These decisions include stopping at traffic lights and crosswalks, avoiding collisions with people and other cars, and finding open parking spaces. After verification, the simulation was validated by observing the utilization over time of each parking lot to ensure it was similar to the utilization we observed from the actual system. The simulation was further validated by observing arrivals to Roswell Road and departures from the parking lots to ensure cars were leaving at the correct times relative to the church service.
Deliverables
In accordance with the results of the traffic simulation model, the following solutions are recommended for PPC to alleviate its traffic congestion:
1) Move service times to create a 30 minute gap between the end of a service and the beginning of the next service.
2) Remove “ghost crosswalks” at the Magnolia Lot and Alberta Drive by removing police assistance at these two locations.
These are the top two recommendations because the simulation resulted in a 23% reduction in Rowell Road traffic from 254 to 196 seconds, a 46% reduction in Rear Surface entrance traffic from 61 to 33 seconds, and a 5% decrease in walking distance when moving service times and consolidating crosswalks were implemented together. This resulted in the highest net benefit for PPC, especially considering no financial cost is necessary to implement these changes.
Bus Schedule
The bus simulation model creates two different kinds of schedules: a four-bus schedule and a three-bus schedule. The four-bus schedule assigns four buses to each of the two offsite lots to operate during the assigned times, while the three-bus schedule assigns three buses to each offsite lot. Different buses vary in capacity. The model identifies busy hours at each parking lot and assigns more buses or less buses depending on the level of crowding. As a result, these schedules are created to reduce the average idle time of buses while maintaining less than 10 minutes of waiting time for the congregants at each lot.
Python Simulator
The Python simulation dashboard will be given to PPC for their future use. It allows a user to simulate the parking system in a simple, user-friendly environment. Users can input dozens of their own parameters or may use the default values. The default values are set to replicate the current attributes of the PPC system according to the data that was collected. This piece of software outputs a set of summary values including: number of cars run through the system, number of cars still in the system, average time a car is in the system, and average time a driver looks for a parking space. The simulation also prints out the occupancy of the parking lots. PPC will have the capability to edit the simulation parameters to predict the results of various scenarios such as a change in congregation size, a change in service times, or the opening or closing of different lots. There is construction planned for 2021 which will close the Gym Lot, cause traffic congestion, and require detours. This is an excellent opportunity for PPC to simulate the construction plans with the Python dashboard.
Project Information
Student Team
Amina Al-Araji, Rian Baxter, Maziar Bezyan, Michael Christian, Freddy Hackney, Christopher Larkins, Nicholas Steinemann, Oneal Suwandi
Faculty Advisor
Faculty Evaluator
