Driver Education project
#Figma #UI/UX
Overview:
I built an interactive AR app on Driver Education with my student design teammates as a solution for eliminating fear related to Advanced driver-assistance systems features (ADAS features). I am responsible for conducting three rounds of user interviews and tests, leading team ideation activities, making the UI, and communicating with developers on requirements.
Team:
A team of 10+ people including Management, Developers, and designers.
My Role:
UI/UX Designer & Researcher since Sept 2019. PM and Design Lead starting from Sept 2020.
Timeline:
Sept 2019 - May 2022
Tools used:
Problem Space
As a member of the student design team University of Waterloo Alternative Fuel Team (UWAFT), which is participating in the EcoCar Mobility Challenge, I took on the challenge with my team to find the solution for Driver Education for Advanced driver-assistance systems features (ADAS features).
The problem is given by the EcoCar Competition as a deliverable to practice design thinking.
Currently, we have seen people holding fear and misunderstanding against the ADAS features, or the so-called AI Features in vehicles. The current introduction process, without proper training and guidance, left users in the midst of trials and errors which causes a significant drop rate in the learning process and danger in future uses.
Hence, the problem statement is as follows:
Develop a solution to educate vehicle users on the ADAS features and eliminate their fear when using.
Quantitative Data:
To validate our assumptions, we also let users rank themselves on a scale of 1-5 for their interest level for automobiles as well as the familiarity for five CAV features. We would use the average score acquired from user surveys to see if our user identification is accurate.
For ACC and AEB:
These two features are more well known in the market.
For “the Experienced”, the average score is 4.61 while for “the Fresh”, the average score is 3.4
For the other less well-known ones:
These other three features ( are less well known in the market.
For “the Experienced”, the average score is 3.9 while for “the Fresh”, the average score is 2.7
The result is: there is a major difference (around 30%) in scores for the familiarity of five CAV features:
First Round Research
Initially, the team focuses on classifying our users further to make more targeted user interview questions. Our assumption is that users who know and have an interest in vehicles would have a different focus and incentive on learning ADAS features than users who do not. There might even be a difference in what ADAS features people would like to learn. Hence, our Research Goal is:
Find what features we should teach our users
Based on people’s interest level
Based on their need for daily driving experience
Validate assumptions that there is a difference in interest level in vehicles among our target users
Validate the assumption that the discrepancy in interest level would cause a difference in needs and interest levels on various ADAS features.
When creating questions, we asked if users have driver’s licenses and if they own a car themselves or practice car sharing. We categorize people with driver’s licenses, owning or having rented a vehicle as “the Experienced“. And the rest is “the Fresh“.
Qualitative Data & Persona:
After validating our users, we went further to interview people in both groups to know more about their thoughts and current pain points. We analyzed the results and document them into personas and used that as a metric later in brainstorming and writing user testing protocol.
The Fresh:
Some of the needs for the Fresh group are:
Learn in a safe way
Learn before getting on the road
Learn the concepts quickly and be able to apply them to all types of vehicles borrowed
Ability to quickly recap
The Experienced:
Would like to use the vehicle to the fullest
Need to understand the all settings and ways to customize for features interested
Need to learn ways to personalize the learning experience
Need to overcome habits from one’s previous driving experience.
Ideation:
Based on these understandings, we started our brainstorming session. We started by individual brainstorming, then organize them using an affinity diagram.
During the session, we managed to carry out three concepts:
An AR/VR simulation with gamification design
Letting people get comfortable with the features before going on the road
Can corporate with car dealers and create educational leisure activities in 4S shops
A set of tutorials implemented in HMI that produces personalized educational learning experience
Video tutorials tailored for user based on their current knowledge level
A full course on ACC which comes with a quiz
Can use driving schools and knowledge test to advertise for ADAS features to new drivers
Idea Testing:
We developed a low fidelity prototype for all three ideas to test users’ interest level.
For a low fidelity prototype, since we have three ideas, our strategy is to “fail fast“. Another reason is to have users give the most honest opinion when presented with a skeleton.
For the VR prototype, we used Powerpoint to mimic what users would see in VR headsets, including the car’s interior, overlay tags, and educational content.
We wanted to put the VR experience at a car dealer shop for people to explore when they are purchasing their first vehicle and plant the seed of interest in an early phase of their driving career.
We used Figma for the AR prototype.
In this prototype, we focused on translating plain words from instruction to more lively and interactive components, such as labeling the car interior with a phone app.
For video tutorial we used Powerpoint to present our ideas to users.
We want to highlight customization part of the learning process by having screening questions and video recommendations based on knowledge level.
We presented our prototypes to potential users for testing. We were delighted to receive feedback like “interesting” and “innovative”. On the other hand, we also found some reasonable concerns when observing users interacting with prototypes, such as:
AR App might be a bit disorganized.
Quotes: “not sure how it relates to the next part of the app”
Too many buttons are not ideal in education.
Quotes: “it would be nice to only use the hand to feel and touch to know which button it is”
Actionable items are placed illogically in the video list.
Would like a start-to-end path when exploring, or at least some guidance.
Testing Result:
VR:
AR:
Video Tutorial:
Prototype Building:
We decided to move forward with AR and VR prototypes as they have proven to be more attractive to our user group.
In this session, we would build a medium fidelity prototype that would be used to perform usability testing to validate the method solves the pain points mentioned before in personas.
AR:
We include more content and organization into the AR app by differentiating the AR feature from other longer learning modules. The user-centered part of the video list is reorganized into a “Feeling Lucky” which lets users learn more about the aspects that they care more about.
VR:
For the VR app, we would like users to mimic more aspects of the VR experience as users said “they might not getting the vibe of VR”. Hence, we let one of our team members sit in the vehicle and join the video call with the front camera on. The interviewee would control that person’s behavior via verbal commands and we would provide hints, just like the audio instruction people hear when interacting with VR technologies.
Validation via Usability Testing:
On average, users are able to complete 10.67/13 tasks for the two prototypes without hints. All tasks can be done with hints. After interacting with the prototype, we also conducted a retention test to validate the effectiveness of the learning result. As a result, all testers can remember key information after relaxing for 5 minutes.
With usability testings of our prototypes, we managed to get further insights into people’s interest levels in both prototypes.
The user expresses more interest in AR apps due to following reasons:
Liked the idea of having real-time instruction and guidance for each button
Worried about the accessibility of VR: “I can’t take that home”
People worried and still had the need to quickly check the modules in the middle of driving
Would like to get quick recaps of knowledge.
“Feeling Lucky” sounds confusing to people
Quotes: “I thought it is a lottery draw or something.”
After we explained the section, users expressed that this would be beneficial to create a more customized learning experience, which is one of the pain points.
Final Solution (High-Fidelity):
In the high fidelity prototype, we chose to assemble an amazing development team to visualize the thoughts through coding in Unity Studio based on Figma screens. This would allow users to actually use our product to learn some knowledge on ADAS features.
The overlay would provide simple information for recalling purposes, addressing the pain point mentioned in testing: The need for a quick recap. If users would like to explore more in detail, they can go into the page for each feature.
The new version of “Feeling lucky”, enables users to explore topics they are interested in by providing a scroll of options
Also, “Tip of the Day” can provide a starter for users who would like to explore like “Wiki-hops”, addressing the need for guidance.
All Screens:
AR Screens:
Based on my interest:
Landing Screens:
I drew some illustrations for landing screens.
Takeaways:
I learned a lot from this three-year project as it spans from the moment I start to learn UX design to now. Looking back the journey is not perfect as I am more knowledgeable in design and UX research; however, it built a foundation for working in a multifunctional team and leading the team in the design aspects.
Since this is a university project, I tried my best to keep the procedure in the best practice. Each round of iteration is a learning process and improving interview protocols, asking the right question, and designing the right features for users are extremely helpful for me to get a well-rounded understanding of UX design.
As the only designer on the team since Fall 2020, I managed to develop leadership skills to lead my team in design-related deliverables such as user-centric design thinking, heuristic evaluations, and information architecture designing. Introducing design to all the people in the team, which is mostly developers, is interesting and beneficial for me to make sure I learned concepts correctly.
Future:
There could be more work to be done, such as conducting another round of usability testing and iterating again. However, this project would come to an end in May 2022 as the EcoCar competition wraps up and team members move on with their post-university lives.