
Ollie
Design a conversational AI assistant for Pittsburgh Regional Transit (PRT)
Project Type
Studio project collaborated with Pittsburgh Regional Transit (PRT)
Team
Yuehui Du (Designer)
Iris Cai (Designer)
Angela Nam (Designer)
Aishwarya Rane (Designer)
Contribution
Value Creation
UI/UX Design
User Research
Motion Design
Prototyping
Timeline
5 Weeks
Sep. - Nov. 2022
Tools
Vioce Flow
Figma
Illustrator
After Effects
Premier Pro


Context
PRT is Pittsburgh's the prominent local public transit system.


Who are the users?

Daily Commuter
who relies on apps to sync bus schedules with routine, aiming to optimize transit and gain recognition.

Bus Driver
who aims to complete ride on time, communicate real-time updates, and assist passengers efficiently.

Disable Passenger
who seeks smooth access and support to enjoy safe, independent ride without feeling like a burden.
Challenge
Passengers rely on multiple platforms but are still unable to communicate with driver and system in real-time to address their needs.

Solution
Interview
What do we want to investigate from the user interview?
Goal 01
Goal 02
Goal 03



Daily Commuter A
Daily Commuter B


Disable Passenger C
Synthesis
Frequent "silent" communication between passengers and the driver

Service reliability is the primary concern for both passengers and the PRT


"Can't be notified instantly" frustrates all target users
Since the "silent communication" only conduct face-to-face, the asynchronous notification frustrates all target users. Passengers can't receive the updates unless get to the bus stop, and drivers can only know accessibility requests when arriving the stop.

Findings
Three major communication painpoints between users and the system
01

Immediate requests increase uncertainty.
Passengers check bus schedule immediately before departure, which leave no room for emergency preparedness.
02

Lack of real-time updates from the driver on-site.
The system hardly get on-site information without driver’s report, neither do passengers.
03

Accessibility support requests can't reach in advance.
Driver can only be noticed the ramp request when actually seeing the disable passenger,
Prioritize target users in different scenarios

Hypothesis of AI
How does conversational AI enhance the communication experience?
By analyzing the benefits of conversational AI and aligning them with identified pain points, we have developed a hypothesis on how Ollie can enhance the communication experience.
Painpoints
Benefits of Conversational AI
Hypothesis

Immediate requests increase uncertainty.

Lack of real-time updates from the driver on-site.

Accessibility support requests can't reach in advance.
Instantaneous response
Conversational AI can respond immediately about the updates and notify about changes.
Effortless interaction through Natural language
Conversational AI enable users to communicate without learn any specific commands.
Text-to-speech with voice synthesis
Conversational AI can enhance accessibility and provide multiple formates of output: text, image, and audio.
Personalized experience through users insights
Conversational AI can provide personalized experience by remembering customer preferences, based on their profile and behavior.
Intelligent Integration across multiple data sources
Conversational AI can integrate with real-time tracking system and multiple users’ inputs to provide accurate information.
Design Direction
From communication painpoints to design directions
Before narrowing down to the specific design, we review the painpoints and digest them to three core design directions to guide the design of Ollie's identity and interfaces.

Define Identity
From the PRT brand identity to Ollie's visual system


Define Ollie's personailty


Storyboard
Explore scenarios about the conversation between users and Ollie

Define Kiosk
Determine the digital bus stop as the secondary touchpoint

Interface Structure
Explore the structure for key features on mobile and kiosk


Visualize Ollie
Visualize Ollie by rapid iterations


Ollie's Motion Status








Final Design
Painpoint 01
Immediate requests increase uncertainty
Solution 01
Pre-schedule &
real-time update notification
Painpoint 02
Lack of real-time updates from diver on-site
Solution 02
Driver provides quick updates via verbal communication
Painpoint 03
Accessibility support requests can't reach in advance
Solution 03
Hands-free request services and notify the driver ahead
Kiosk Feature 01
Personalized ride recommendation
Kiosk Feature 02
Cross-platform information synchronization
Kiosk Feature 03
Real-time notification broadcasting
Scenario 01. Daily commuter schedules a ride



Scenario 02. Driver updates bus status to passengers at the next stop



Scenario 03. Disable passenger requests support to the incoming driver



Scenario 04. Passenger asks for a ride from kiosk and export to the mobile


Scenario 05. Kiosk broadcasts bus updates to the passengers waiting in line


What I've learned?
Emerging technologies are not the answer, user's needs are.

Next steps
Emerging technologies are not the answer, user's needs are.

© Yuehui Du 2024