Carriage provides mobile and web platforms that improve the scheduling, performing and monitoring experiences for CULift, Cornell's transit service for students with disabilities. Carriage Admin enhances the scheduling and monitoring of student ride requests for CULift administrators.
I led design and implemented the data analytics, platform navigation, and live status features. I discovered the problem, defined the focus, developed potential solutions, and delivered accurate results through designing, prototyping, and user testing.
In the current system administrators have to manually input rider and driver data on an excel sheet from which it's not only hard to gather insights but also near impossible to detect when something goes wrong. Furthermore, navigating through different dates and compiling necessary data is extremely tedious and time consuming. Thus, the main problem is
"How might we analyze ride data and discover anomalies more efficiently?"
It is important to note that within our product this problem applies to a niche group of people. So the first step in solving the problem is to understand how these users interact with the current product and what their pain points are. The big questions I will be asking are:
1. What are the common characteristics in the users?
2. What is the user flow in the existing product?
3. What are the pain points in the current user flow?
The administrator's main responsibilities are to schedule rides and dispatch drivers, so they want to quickly notice problems with rides, find its the root cause, and come up with a solution.
Behaviors
Frustrations
Goals
From insights I started exploring various interactions. Resolving the most time consuming pain point - browsing existing data, meant implementing an analytics feature that organizes the complex data set and makes it easy to pin point a specific group of data. After user testing, the interval picker and summary dashboard especially appealed to the administrator because it helped filter the unnecessary data and visualize the necessary data.
The analytics feature only solved part of the problem because there was no precise ride feedback system that allowed for solving today's problems in a timely manner. So, I implemented a live status feature that would provide up to date information on today's rides and highlight any anomalies.
To ensure that the platform is accessible and to reduce complexity I wanted to use soft primary colors with enough contrast to be easily visible. I also implemented distinct icons for easy status recognition.
The administrator Mindy wants to find and solve any problems with rides. Let's visualize her journey.
Before Mindy starts to recognize any problems, she needs to go to today's schedule.
Schedule Navigation
Mindy then needs to delve a little deeper into the problem rides to see the exact issue.
Live Status
Mindy wants to understand the root cause of the problems, so she uses the analytics feature. The data is filtered yearly and monthly or by using an interval picker. The aggregate is then shown in the dashboard.
Analytics Feature
Mindy remembers to check for changes to existing schedules to see whether that is the real cause instead.
Notifications
Besides working on navigation, analytics, live status, and notifications, I was also helped develop the features for giving new admins access to the platform. Below are renderings of features I worked on that are not listed in the case study.
