04
IntelliRide
How might we integrate physical and digital controls for the semi-autonomous electric bike to ensure a seamless commuting experience?
My Contribution
User Research / Task Analysis / Product Design / UX Design / Prototyping / Usability Testing
Project Type
School Project @ Carnegie Mellon University
Team
2 Product Designer
Timeline
7 weeks
Background:
Future Semi-Auto Electric Bicycle
IntelliRide is a conceptual semi-autonomous electric bike featuring Level 3 driving automation. In this project, we mainly focused on the interaction between users and autonomous systems, including how users control and perceive information on the bike, both physically and on screens.
Design Overview
This e-bike’s autonomous system could set routes, navigate, and make decisions based on real-time conditions. Users can delegate control for most of the journey, intervening only in emergencies.
Flexible Control in Autonomous Mode
Riders can choose the amount of control they would like to have at any point during the autonomous mode.
Intuitive Controls
The integrated physical controls make adjustments more intuitive and safer.
Smart Route Optimization
Riders can choose route preferences in the phone app to reflect their riding priorities. The app also allows users to provide feedback to improve future routing.
Stay Connected and Safe
The display will show important phone notifications to the user when their phone is connected to the bike. The user can respond to those notifications with voice control.
User Research:
Exploring bikers’ needs
At the beginning of the project, we conducted guerrilla research and interviewed several bikers around the campus to understand their pain points on biking. We also observed e-bike parking on the streets and visited a bike shop in Pittsburgh to learn more about the existing e-bike design and market. Here are the user needs we found:
Accept the New System
People are generally hesitant about giving full control to the autonomous system, thinking it might not be “smart” enough to deal with every situation during the ride.
Customized Route Planning
Many of our interviewees mentioned that they would choose routes based on traffic conditions, time, and safety. However, most bike navigation systems don’t cater to how users plan their routes.
Easier Control
Many commuters use e-bikes to commute because they are faster and effortless. However, many of them find e-bike controls complex and overwhelming.
Staying Connected while Riding
Many e-bikes are equipped with two screens (the dashboard and phone for navigation), meaning people still use phones when riding for route searching, navigation, or texting.
How might we make the autonomous system more acceptable to users?
How might we reduce cognitive load for users while riding?
How might we facilitate an efficient commute for users?
Challenge 01
How might we make the autonomous system more acceptable to users?
Design Strategy:
Autonomous on a Spectrum
We assumed the autonomous system would control the steering according to the route it set and adjust speed based on the surroundings and the actions (e.g., go at the green light, stop at the red light, and slow down when turning).
In our approach to autonomy, we aimed to empower users with a sense of control even in fully autonomous mode. By defining individual spectrums for crucial bike operations such as acceleration, steering, and braking, we established a dynamic system where the autonomous influence adjusts based on user inputs. This allows users to override specific aspects of autonomous operation at any given moment, ensuring a seamless and flexible riding experience.
Design Solution:
Flexible Autonomous Mode
1. Activate Different Modes of Autonomous
When the users press the “Autonomous” button, the system will ask if they want the Autonomous mode to navigate for them or just control their speed.
If the users choose “Autopilot,” they can select destinations from the “starred route” to start moving immediately or search for their destination.
On the other hand, if the users choose “Speed Control Only,” the system will only control the speed, and the users will control the direction.
2. Take Over and Override Autonomous Mode
When the system detects something happening and needing the users’ control, a take-over message will appear and show instructions that tell users what to do to override the system seamlessly, preventing possible accidents caused by users’ nervousness.
Challenge 02
How might we reduce cognitive load for users while riding?
Design Strategy:
Simplify Displays and Controls
From the interview with staff from a bike shop, we learned that e-bike controls can be overwhelming for users due to their complexity. Therefore, it is essential to (1) present bike status information intuitively to enhance user understanding and (2) organize controls based on ergonomic principles, facilitating easier adjustments for users.
Design Solution:
Intuitive Design
1. Dashboard Design
Since we put the screen on the lower part of the headset to avoid blocking the user’s view, we decided to place the most crucial information (e.g., speed, battery status, autonomous status) in the top bar of the dashboard so that the user can skim the information easily when riding.
We use a white border to indicate manual mode and a green border to signify autonomous mode, making it easier for users to identify their current mode.
Users can switch between the homepage and map view depending on how much information they want to see during the ride.
2. Physical Control Design
We aimed to keep all the buttons within a thumb’s length of either hand when riding to reduce the need for users to remove their hands from the grip before reaching for any controls.
Left
1. Blinker
2. Autonomous Button: As the main feature of the semi-autonomous bike, the “Autonomous” button is large so that users can easily find it.
Right
3. Assistance level
4. Trackpad: We designed a trackpad that enables users to navigate the touchscreen without lifting their hands from the grip. The cursor would snap to clickable elements for easier control.
Grip haptic: Whenever the Autonomous system is about to take action, haptics in the grips will vibrate in sequence to communicate that to the rider.
Design Iteration:
Testing the Riding Experience with Prototypes
1. Dashboard Design
Paper Prototype
For the first round of testing, we built a physical prototype with paper and cardboard to validate our concepts of the whole riding process. We installed our paper prototype on an office chair to move the participants when they activated the autonomous mode, simulating the riding experience with the autopilot system.
Digital Prototype
Once we validated the user flow with the paper prototype, we refined our design and created a high-fidelity digital prototype for user interaction testing. Using the Wizard of Oz method, we simulated both autonomous mode and scenarios requiring users to interact with the digital dashboard using physical controls.
2. Physical Control Design
Distributing the Controls
To enhance user control while riding, we planned control placement to ensure accessibility while gripping the handlebars. We began by sketching a diagram (image 1), organizing controls into logical groupings on either side. Then, we built a paper prototype (image 2) for testing user comprehension without extra guidance.
Iterating Control Size
During the usability testing, we realized that the size of the controls was too large for people with smaller hands (image 3). Subsequently, we iterated on the control size, refining them through sketches (image 4) and clay models (image 5) to enhance the ergonomic design.
Challenge 03
How might we facilitate an efficient commute for users?
Design Strategy:
Efficient Commute with Safety Focus
Recognizing the significance of time for commuter bikers, we incorporated features aimed at utilizing their commute time while prioritizing safety. The system would collect user preferences and feedback to optimize the route recommendations. It also allowed users to respond to messages with voice input during the ride, utilizing their time while preventing them from being visually distracted.
Design Solution:
Adapting Routes Based on Shared Experiences
Before the ride, users can set up route preferences (e.g., shortest time, most bike lanes, best scenery) in the phone app to reflect their riding priorities. The system will find routes based on these settings.
After the ride, the App would save the route’s information (e.g., distance, riding time, and battery level) as training data. Users can add the route to “starred routes” for quicker access from the dashboard.
After each ride, the system would ask for the user’s feedback on the ride experience, like “Why did you turn left on S Craig St?” to optimize the route recommendation system. We also implement community-sourced navigation routing by aggregating paths chosen by riders with similar route preferences.
Design Solution:
Stay Connected to the World
When users receive notifications on their phones, the bike will display them on the screen and provide the option to respond using voice commands.
If users have an upcoming calendar event with a location or receive a text message containing an address, the bike will temporarily add those destinations to their starred locations for easy access.