We created a dashboard interface that conveyed important information and the status of certain features, but still felt familiar to the user. The user can easily get directions to their destination and follow them, or use a new feature, autonomous mode, which will do the navigating for them. We especially focused on the battery and how it would impact the overall user experience, since it is an integral part of electronic bikes. If the bike has low battery, the user is warned when the battery will not last through certain routes.
Semi-autonomous E-bike
There is a paradigm shift towards semi-autonomy, and while cars are at the forefront of this change, other forms of transportation need to catch up
The Problem Space
The future is branching away from our car-centric past—and businesses need to adapt. The Ford Motor Company wants to develop semi-autonomous electric bicycles in response. These will utilize level 3 autonomy, meaning humans do not need to pay attention to the road, but they do need to be ready to take control, if required. My team was tasked with designing this new form of e-bike, focusing on digital and physical controls.
How might we convey new concepts like semi-autonomy, while still matching users' mental models of e-bikes, so that they can use our design seamlessly?
The Solution
Autonomous Mode & Pedal Assist
We decided that pedaling should be decoupled with autonomous navigation, to give users more control over their biking experience. Pedal assist moves your bike for you, with an option for half assist or full assist. Autonomous mode navigates for you, balancing the bike in the process. Seperating these two modes suits the needs of a larger demographic of people, from the exercise-savvy, to the modern office worker.
Take-over Moment
Since the e-bike is semi-autonomous, there may be times the user needs to take over control of the bike. We implemented pop-up screens meant to alert the user without being too panic-inducing. If the user does not take control, the bike will pull over and instruct the user to prepare to put their feet on the ground.
In addition, we designed an app that allows users to feel connected to their bike even when they are away from it. The features reflect this focus--the locate tab shows them where their bike is, and can give directions in case the user forgot where they parked their bike. The security feature gives live camera feed to the user. Other tabs give stats on the bike, or the health of the user.
Bike Location Finder
Users might not always be near their bike, or it could be in a location they are not familiar with. We designed a feature that locates the bike and gives users directions to it.
Security
Security was a major worry of e-bike users that we uncovered during guerrilla research. We designed a camera monitoring system so users can watch their bike from afar.
Design Process
We started with guerrilla research and semi-structured interviews to gain a better understanding of the problem space and pain points of current e-bikes. People used e-bikes for a variety of different things—for making deliveries, in place of cars, or for exercising when they no longer had the health to ride a regular bike. Thus, our new design needed to be suited for a wide variety of needs.
Next, we constructed a task analysis for when users use e-bikes. This would help us look at the big-picture and determine how our design could influence different steps
We decided to take aspects from planning the ride and integrate them into the e-bike. Instead of the user taking out their phone, pulling directions to their destination up, and then looking down and squinting at their phone while trying to follow the directions, we could display them on a large dashboard interface. All of these steps are simplified, and the directions are easier to follow during biking, which requires most of the user's attention.
Arguably the most important step in the task analysis is riding the bike, everything else is in preparation for it. It also imparts the highest cognitive load. Therefore, introducing semi-autonomy to this part of the e-bike experience could prove extremely useful. What if an UberEats driver could put in an address, hop on their bike, and be autonomously taken to the location? While some users would love this option, other users use e-bikes for exercise. What if they don't want to think about navigating, but still want a workout? We decided to separate "navigation" from "movement", making them two separate features called "auto mode" and "pedal assist". This gives users more flexibility in their biking experience, and can appeal to a wider ranger of users.
A major problem we faced was communicating our novel design solutions with the user. The wifi symbol is now ubiquitous, but when it was first conceived, how did people know what it was? The same goes for the play button on a controller, how does a triangle convey "start"? We had to navigate these unknowns in our project with auto mode and pedal assist. We created little introductions for each mode, which would appear when the user uses the bike for the very first time. These tooltips can be accessed via the "i" symbol in the bottom left corner.
One of the most illuminating steps in the design process was usability testing. We conducted many usability tests on our low fidelity and high fidelity prototypes to learn key pain points or misunderstandings. Our takeover notification screen went through many iterations because of the feedback of participants. At first, we made it take up the entire page, which users said was too scary. Then, we put it as a pop-up in the corner, but users said it did not seem urgent enough for the nature of the situation. Participants also suggested icons or images of what they needed to do, instead of only text. Finally, participants wanted to know how much time they had before the autonomous mode turned off. All of these suggestions were critical to shaping our final product.
Another point of confusion was the difference between auto mode and pedal assist. Before we added the information pop-ups, users could not figure out how the two modes differed. The language we used for the modes in earlier iterations was confusing, as well. We went through many different icon designs, display concepts, and names for the modes before arriving at our final product.
