Presence

​Helping family members travel safely and independently while feeling connected to their loved ones. Designed at Stanford for Hyundai Motor Company.

Skills
User research
Physical prototyping

Timeline
Jan - Jun 2021
20 weeks

Collaborators
Mechanical Engineers: Crystal Winston, Frances Silva Roig, Nathan Poczekay

Software Engineers: Ben Hurdelhey, Carla Terboven, Hendrik Patzlaff

Overview

Presence is a communication system for fully autonomous vehicles. It enables each family member to travel safely and independently, while feeling connected to their loved ones.

​Presence consists of​ (a) a Face-ID camera module that recognizes who is in the car and sends multiple camera perspectives to the app, (b) a transparent video calling screen, and (c) a mobile app for family members outside of the car.​

Defining the Problem

Hyundai asked us to come up with a communication system for autonomous cars, 10-15 years into the future.

From initial needfinding interviews and research, we learned that it is a huge burden on parents to manage family transportation logistics:

We decided to explore how autonomous cars could enable older children (ages 10-16) to transport themselves independently, saving parents significant time and stress.

User Research Findings

While developing Presence, we engaged over 60 parents, children, and caretakers in order to understand what families would want before adopting independent transportation.

Parents need to be assured that their child is safe.

Parents will only use Presence if it is completely fail-safe. Safety concerns fall into two categories: uncertainty about the safety of autonomous cars, and concerns about leaving their child without direct adult supervision.
Families do not want to miss out on quality time.

​A video or phone call is not a good enough replacement for being in-person. Parents and children want a more seamless and immersive way to communicate remotely.

Designing the Parent Experience

We started by designing for the parent's perspective.

To define what features would satisfy parents' needs, we made a series of "Wizard-of-Oz" prototypes. These prototypes enable users to experience what a system would be like before it is fully built.

Camera Views Prototype

How can we make remote video calls feel more engaging and seamless, like being in the same room?

We had 6 participants chat with a car passenger over zoom while giving them additional camera views of the car's interior and exterior.

Zoom call with additional camera views of the car's interior ("Front Inside") and where it is going ("Front Outside")

Participants liked the Front Outside view and said it "felt like being together". They could see where the passenger was going and many chatted about the car's surroundings. This influenced the final design:

  • By default, include the Front Outside view on the calling screen
  • Make the Front Outside view wide-angle, so the outside caller can see even more of the passenger's perspective

Concept Video Prototype

How can we get concrete feedback from parents, without triggering fears caused by the current underdevelopment of autonomous vehicles?

A video prototype of our original concept, called "Car Butler"

​Most parents we interviewed were uncomfortable with the theoretical idea of having their kids in autonomous cars.

This video made it easier to imagine the concept in the real world and gave parents concrete features to discuss.

​We received feedback on this video from 12 parents and 2 kids in middle school. Parents suggested additional features and asked:

  • How can my child be sure they are entering the right car?
  • What happens if a stranger attempts to enter the vehicle with them?

Our final design requires two-factor authentication for operating the car: a key to enter, and FaceID to make sure only recognized passengers are in the car. The app also includes a wide-angle view of the car's interior, so parents can see exactly who is in the car.

Final Parent Experience

Our final app incorporates the additional camera views and location tracking that we prototyped, alongside the ability to call particular contacts. I used our research findings to give design specs to my software engineering teammates, who implemented the app.

See who of your family and friends are available and choose who to call

When calling a passenger, you can see their surroundings and feel like you are together

Zoom out to see multiple passengers, the ride status, and the car’s location

Designing the Passenger Experience

Once we validated that parents were interested in a system like Presence, we started to explore the passenger's experience inside the car.

We learned early on that opaque screens contribute to motion sickness, so we focused our experiments on transparent/translucent screens.

Screen Prototypes

What should the screen's material, size, and position be? What other factors contribute to immersive communication?

For this experiment, we adhered projection film onto acrylic and projected a video call onto the transparent surface. Then we had participants use the "screen" to chat with someone in a different room.

We learned that removing the caller’s background makes conversation more immersive because it feels like they are physically present.

Prototyping video calls on a translucent screen. Participants liked the empty background (left)

We also had users participate in video calls inside a car, in order to define a screen location and image size that would feel the most natural.

Testing different screen locations and image sizes

Some key takeaways:

  • Screen MaterialCompletely transparent screens are not ideal because it makes it more difficult to focus on the caller's face. Our final projector film has a slight opacity that blurs the background.
  • Image SizePassengers find it more engaging when the caller’s image is life-size. However, if the screen is closer than their conversation partner would be in real-life, the image should be scaled down proportionally to feel natural.

Final Passenger Experience

​Our research and tests helped us define detailed requirements for the implementation of Presence. We built out these features for our final prototype, so that people could experience what Presence would be like if it were fully implemented:

TRANSPARENT SCREEN
​makes it seem like the caller is physically present in the car and reduces motion sickness

LIFE-SIZE IMAGE
​creates a more immersive conversation experience for the passenger

FACE-ID CAMERA MODULE
​recognizes who is in the car and provides multiple camera perspectives to the app

ADJUSTABLE DESIGN
​accommodates passengers of all sizes with a rotatable screen that folds down from the ceiling

Moving Forward

We have received a lot of positive user feedback on Presence. The safety and security features help parents feel more comfortable about letting their children ride independently, and people find the communication system more engaging and immersive than a normal video call. In order to make Presence a more complete system, our next step would involve enabling families to handle ride logistics and scheduling through the web application.

We've now handed off this project to Hyundai​'s User Experience Strategy Team. Our complete documentation can be viewed here, and includes more information on our secondary research, competitive analysis, and design requirements.

Reflection

I learned a lot from this extended, generative project. At each step, I got to practice identifying the critical questions we needed to answer and deciding how to answer them — whether through online research, an internal discussion, or running an experiment.

I also learned to be aware of my own cultural context and assumptions. Half of my teammates were located in Germany, where it is common for young children to travel on their own via public transit. ​In countries like Germany, Presence would be much more relevant as an immersive communication system for commuters than as a safe, convenient form of transportation for kids. We tried to keep both contexts in mind while developing Presence.