MBUS - Integrating Route Prediction Model Transportation App
Course Project
Back-end Development
Prototype
My Role
Developer
UI/UX Designer
Area
Route Optimization
Prototyping
API Development
User Interface
Tools
Python
Figma
Teamwork
Joshua Nicolson
Devyani Jain
Current issues in the MBUS SYSTEM
Bus routes lacks adaptability and fluidity
Doesn’t reflect real-time predictions. Just a permanent schedule. eg: construction, snow and traffic delays.
Hard to access frequently used routes
Doesn’t reflect real-time predictions. Just a permanent schedule. eg: construction, snow and traffic delays.
Language barriers
Doesn’t reflect real-time predictions. Just a permanent schedule. eg: construction, snow and traffic delays.
Unreliable route timings and frustrations
Doesn’t reflect real-time predictions. Just a permanent schedule. eg: construction, snow and traffic delays.
The SOLUTION
Timeline
Timeline
Phase I: Developing the Route-Prediction Model
USER Data structuring and API integration
The Ride API has unorganized data
Python-based backend fetched and organized data using:
Dictionary that store bus stop IDs as key and arrival time as values
Nested List that presents sequential data for bus routes
Impact: Organized data allowed faster access and a clear relationship between user-selected stops and arrival times
Handling user input
The system required users to enter a six-digit stop ID to search a stop!
A dictionary that
mapped bus stop names to their stop IDs
Impact: Now, users can look up a stop by it’s name and not the bus stop ID number
Error Management
The API failed to return data for specific requests due to delays, missing information, or inconsistencies in route schedules
Solution: Add error handling parts
Validating API data before display
Providing alternative feedback or default messages
Impact: Minimized the risk to displaying inaccurate data
The class was over at this point, but
"What would this back-end system look like if it was implemented visually?"
Phase II: Designing the Interface
Updating the Challenge
How might I design re-design the M-Bus application implementing the route prediction model to improve the experience?
User Interviews
User Interviews
After looking at The Ride and current M-Bus app
I interviewed 10 Sophomore students at The University of Michigan from different majors.
I asked them about their current experiences with the Bus system and tracking it down.
Feature will be implemented?
Persona & journey
An Overview of a User's Journey on the Bus
Both, the persona and journey mapping helped inform pain points, which areas to be focused and how the current prediction model can be impactful.
Navigating multiple users; student, professor and faculty & Empathy Mapping
Assessing the old UI
What can be improved?
Comparison key takeaways
Visualizing the route prediction model
Low-fidelity wireframes
I iterated over these wireframes a couple of times. Looking back would have conducted usability testing to iterate further :(
Integrated Route-Prediction Model
M-Bus Transportation App
Catch your bus, not the stress. The U-M Bus App gives you real-time updates and instant access to your favorite routes, so you’re always on time
Expanding Impact
Persona & journey
TAKEAWAYS
Next Project
Inclusive and Accessible XR on VisionOS Application
/Design Lead
/Project Management