Team Name
Team 3.14159
Timeline
Fall 2018 – Spring 2019
Students
- Jacob Devasier
- Ethan Duff
- Seth Tisbi
- Kevin Tiller
- Miguel Fraire
Abstract
Traffic Pi will be able to analyze the speed of individual vehicles entering its view area and record this information to a centralized database. The expected use of Traffic Pi is for its user to direct the camera at the area of intent, turn on the device, and begin the traffic study on the targeted area.
Background
The current method by which a traffic study takes place can be generalized by the following. A need for such a study arises in a neighborhood (multiple accidents over a period of time at a particular intersection, multiple close-calls at a particular intersection, etc.). Either the city recognizes a potential problem in the area of interest or a resident in the community makes known to the city the area and requests something is done. The city will usually follow up on the request with a traffic study; monitoring the area with pneumatic road tubes, wires laid on the ground that monitor the speed of the vehicle as it passes over it. Depending on the area of interest, multiple road tubes may be used along with other equipment or actual persons on the ground monitoring the area. After the area has been monitored for a given period of time a decision is made on the area (install stop sign, signal lights, road bump, etc.). There are several problems with the current model.
To begin, the ultimate decision to initiate a traffic study begins with the city. A citizen, neighborhood,
or even business park may request/petition the city to investigate potentially dangerous area but these
are only requests. If in fiscally troubling times the funds for such a study are lacking, the study may not
even be feasible and thus not occur. Additionally, if the city for whatever reason (politically motivated,
not enough complaints, etc.) decides not to investigate the area, the public is not able to have their request satisfied.
If a city does decide to begin a traffic study, the cost of such a study may limit the scope or duration
of the study. On average the cost of a single pneumatic road tube can range from $500 to the $1000’s. This cost does not include the installation cost, environment hazards, maintenance, and otherwise. This cost is heavy on both large and small cities. In large cities, there is likely an abundance of areas that require a traffic study and thus multiple devices must be deployed. In small cities, the budget for such a device may not exist at all. In short, the cost of the current method is high.
Lastly, in relating to the above two points, the ability to perform the study lies solely within the city
because of power and money. The individual citizen, neighborhood, or entity is notable due to the
reasons above able to conduct this study on their own. Relegating what should be a local affair to city
governance.
Project Requirements
- Find the vehicle passing.
- Track the passing vehicle.
- Predict and calculate the distance of the car from the Traffic Pi.
- Calculate the speed of passing vehicles.
- Memory storage in a database.
- Preprocessing of the data on the pi.
- Post-processing of the data on local machines.
- Cost and affordability of the Traffic Pi.
- Visual representation of data.
- Mountable and stability of the Traffic Pi.
System Overview
The architecture revolves primarily around the raspberry pi computer itself. The interface of the raspberry pi will communicate with all other major systems sending commands and receiving data. The first major system to communicate with is the vision system. This system is responsible for the cameras and has two that are part of it: the IR camera, and the 3D camera. Both will connect to the PI’s interface in order to transmit visual data. This data will be redirected by the interface to the Pre-processing unit which will determine what raw film data is worth keeping and what is not by a factor of cars in the streets. Data kept is placed into storage which the interface can then pull out again. It is powered by a socket connected to the power system, an outlet inside the users home. The case system has a cooling unit which prevents the PI from overheating through a ventilation shaft. This fan draws power from the PI via its interface. When the PI is connected to the user’s computer via Ethernet, it will automatically run its application on their computer. This application receives transferred data from the PI’s storage and processes it, splitting metadata into a database and film data into local storage. All of this data will be accessible to the user via a graphic user interface.
Results
Future Work
- Compatibility with Cloud System
- Pi Network
Project Files
Project Charter (link)
System Requirements Specification (link)
Architectural Design Specification (link)
Detailed Design Specification (link)
Poster (link)
References
Any references go here, properly formatted
