Inspiration
The inspiration behind CampusETA stemmed from the collective frustration and inconvenience experienced by Duke’s student body and faculty using the existing TransLoc app. While TransLoc provides basic bus location information, it fails to deliver a streamlined user experience, especially when planning commutes around campus. Considering the universal experience of inconsistent bus routes, such as the C1 and Swift lines, we recognized a critical area for improvement in daily campus life, especially among first-year students who rely on Duke transit daily. We aimed to enhance the commute experience by offering precise bus arrival and destination time estimates in a more convenient and accessible user experience, directly addressing the need for a more reliable, user-friendly transportation tool.
What it does
CampusETA provides an upgraded tracking service for the Duke bus system. Unlike TransLoc, which only tracks bus locations and provides inaccurate arrival times, CampusETA streamlines the user experience by providing accurate bus arrival times, predicts time to destination, forecasts how busy the buses are, and features a widget that users can add to their lock screens for added convenience. Users can quickly glance at their phones to get real-time updates without needing to unlock their devices or navigate the app, further increasing the efficiency of CampusETA.
How we built it
Our team developed CampusETA using a combination of GPS data, real-time API calls, predictive algorithms, and user-centered design principles. We integrated the real-time data from the existing bus tracking systems and applied predictive modeling to calculate accurate arrival times, taking into account variables such as historical travel times and time of day. The app was developed on Swift, but we ran algorithms in Python. We paid particular attention to the widget creator in Xcode, ensuring it was intuitive and informative, making important information accessible at a glance.
Challenges we ran into
One of the main challenges we faced was dynamically optimizing the UI, given the user’s current location. CampusETA uses the user’s GPS data to find the closest bus stops and potential routes they might take. However, in some edge cases, such as by the Freeman Center, multiple stops with overlapping destinations such as East, West, or Swift exist. To cover these, we developed an algorithm to cluster nearby stops into a hub, which can route to any destination. Another challenge we faced was optimizing API calls to avoid ratelimiting the API. We plan to move this to a cloud server to scale our potential deployments.
Accomplishments that we're proud of
We are proud of creating a user-friendly app that addresses a relevant transportation need within the Duke community. The successful implementation of the lock screen widget demonstrated a successful implementation of our initial goal to provide an easy-to-use, easy-to-understand experience for all users. Our ability to overcome the technical and algorithmic challenges to provide accurate, real-time bus tracking and predictions marks a significant achievement. From our initial user experience testing, we saw reasonable loading times and easy-to-understand information from the app, which we hope incoming first-year students will find convenient.
What we learned
Throughout this project, we learned the importance of generalizing algorithms for a variety of edge cases. Coming into the project, we naively assumed the simplest bus route with two destinations. However, the reality of complex overlapping bus routes required the complete overhaul of naive algorithms for adaptable and universal procedures. We also learned the importance of user-centric design, going to far lengths to ensure ease of use and accessibility. Rather than attempting to improve Transloc’s data, our project relied on improving the user experience and emphasizing the actual needs of the user.
What's next for CampusETA CampusETA uses data for Duke bus routes: C1: East-West, SWS: Swift Avenue Shuttle, and CSF: C1-Swift Avenue. In the future, CampusETA will provide data to all on-campus transportation routes and allow users to select preferences for stops, routes, and departure reminders, providing a personalized user experience. Furthermore, the data we used to develop CampusETA applies to other university campuses and could incorporate a wider variety of uses, such as comparative measures for different modes of transportation or recommendations for the fastest arrival times. Additionally, we will make our future representation for the busyness of each hour specific to each route while being quickly navigable for a user.
Log in or sign up for Devpost to join the conversation.