The goal of our project for senior design is to develop cell phone applications using BREW (Binary Runtime Environment for Wireless).  This past year we worked on developing 3 applications; CellCard, ParkingTrak, and Edge Detection software using Matlab.  The goal of our project is to design and implement an efficient program that will be easy to use, and memory efficient on the cell phone.  We had many difficulties in the beginning of our project, because of the lack of books, and support on the web. 

Our first application is CellCard.  The application’s main goal is to display a barcode on the cell phone screen, to expedite credit card transactions.  The barcode displayed will be compatible with current scanners.  Credit card use is substantially increasing, as interest rates drop.  Cell phone use is also increasing substantially, due to improvements in technology, and price reductions.  BREW, developed by Qualcomm, teamed up with Verizon Wireless to provide an environment coding based on C, to be used for cell phone application development.  We have two prototypes for this program, which will be explained later on, that hopefully can be improved upon next year by other group members.  Some additions and improvements we have made to this code is that now the code is more efficient, and the total length of the code has dropped significantly.  Originally, when we first started programming, we had an extremely difficult time understanding how BREW works; because although based on C, the lack of books and instructions impeded our progress.  Discussed later in this paper, the paper will explain how we overcame such problems.

Second application that we developed was ParkingTrak.  This program we feel is the integral part of our learning experience using BREW.  Originally we were planning on developing just the CellCard application, but as the completion of the program neared, the group felt that our initial program was inadequate for our final project.  ParkingTrak is a program that tells you how many spots are located in each parking lot at Stevens Institute of Technology.  This code basically communicates with the server, and downloads the necessary information that is needed to let the user know how many spots are available.  This program can be implemented at Stevens with almost no costs, because everything is software based, and the user pays a one time fee to download the application. Our ParkingTrak software is unique because it uses Edge detection to tell you exactly what lane and how many spots are free. Some improvements that can be made to ParkingTrak is to maybe display the map of the parking lot before it tells you how many spots are available.  

The final application that we completed developing is Edge Detection for ParkingTrak.  This software basically works in conjunction with ParkingTrak to update the server using different database software.  Edge Detection is developed using Matlab.  3 different prototypes were discussed before implementing edge detection, which will be explained in the report.  Edge Detection basically takes the intensified image of the parking lot, and calculates the mean of the white edges.  This will tell you the amount of cars located in each lane.   Improvements we made from our originally prototype is that it can successfully count the number of cars in each lane.  We hope maybe another group can take over our project and actually implement the database side, so the server can be updated automatically every 5 minutes, by taking snapshots using Edge Detection.

These 3 applications developed by us, we feel that is adequate enough to meet the senior design requirements.  We learned a great deal about how BREW works, and hopefully in the future we can develop more applications using BREW.  The rest of the report will discuss the various tests, and prototypes used to help achieve our goal.  With help from our advisor Professor Yao, we have successfully completed the 3 goals of our project to make.