Boston’s Green Line subway riders know well the futility in following the system’s published schedule; unless you are catching a train at its origin, you won’t see it arrive “on time.” Unfortunately, much of the Green Line runs above ground on the same streets as automobile traffic, so its trains are subject to the same stop and go rules of the road. Add in the need to wait for large groups of people to exit and enter the trains at each stop and what you get is a painfully unreliable commute.
ECE Senior Design team “Macrosoft” (Abin Ajayakumar, Moses Chen, Nael Musleh, Christopher Ogorzalek, and Patrick Raspante) designed an innovative system called “T-NAV” that aims to at least partially solve this problem by adding a measure of predictability to the subway waiting game.
The project is based on a collaborative effort between GPS-enabled cell phone users to create a network of nodes feeding data to a central system. When the T-NAV cell phone application is activated by a user while on a train, the phone will send anonymous position data to a centralized server. The data is then run through a series of calculations and compared to current data and historical data from other users. The result is a system that can predict where the train currently is, and at what times it will arrive at the remaining stations in its route. End-users can then access this information on a public website.
Learning the intricacies of cell phone systems and programming for mobile devices were noted as major hurdles for the project; no one had experience in those areas. But it was the “nitty gritty” that really tested the team’s ability to successfully complete the system.
“Since the project consisted of many sub-components that were being developed in parallel, it was particularly hard to make it all work together,” said Ogorzalek. “Everyone knew that the project was ‘almost done,’ and it was that way for a good three to four months before we could actually demonstrate full functionality. Small things like a website pop-up window grabbing data from the central server in the wrong way, or a minor bug in the train motion logic, created cascading problems. Resolving these issues required a lot of testing and debugging time and a lot of patience with teammates. Communication and cooperation was key.”