Moving goods to moving people—transferring the Aurora Driver from Class 8 trucks to passenger cars.
Aurora Connect
Going beyond freight, the Aurora Driver is built on a common architecture capable of powering a wide range of vehicle types. We demonstrated this flexibility with Toyota on the Sienna platform. As the Aurora Driver unlocks the benefits of self-driving for both roads and the broader supply chain, Aurora aims to extend those benefits well beyond freight.
To test and validate this strategy, we assembled a small, agile team drawn from product, engineering, design, and our partners at Toyota and Denso. This initiative was driven by the voices of our future customers.
Design, engineer and produce a fleet of roadworthy autonomous vehicles intended for ride hailing application
-
First external user research to bring user voice into the autonomous vehicle development.
-
Fostered co-ownership cross functionally through design-led initiatives.
-
Design to production of small fleet of operational vehicles in 2023.
-
Physical touchpoints needed to deliver a safe and comfortable rider experience.
-
-
Provided successful demonstration rides to Uber and Toyota executives on autonomous rides in Texas.
-
Hardened partnerships and plan commits from Uber and Toyota
-
Built internal muscles on corss functional development processes.
-
Chris Matsuda, Sarah Mazzochi, Dan Kanitz, Andrew Moore, Horatio Cowan, Uber team, Toyota team and many more.
Fall 2023
In 2023, we had a unique opportunity to engage with and learn from the very first riders of self-driving cars.
Market Research
Benchmarking — is most valuable when it reveals gaps in competitor offerings giving us a clear opportunity to focus and outperform in those areas.
Research Premise
2023 was an inflection point for the self-driving industry. Cruise launched its ride-hailing service to the public in San Francisco and Austin, while Waymo expanded access to select members in San Francisco and Phoenix. This gave us the opportunity to speak with some of the few early riders who experienced self-driving cars for the first time.
During a series of 60-minute interviews, we used four structured exercises and discussion tools to surface riders' stories, needs, hopes, and concerns. These conversations centered around five key moments across the rider journey—spanning both physical and digital touchpoints.
Overarching questions
-
How are riders responding to AV rides offered today?
-
What are some initial expectations being set?
-
Where are riders experiencing friction and delight?
Key areas of focus
-
Vehicle identification and pickup
-
In-vehicle amenities and cabin experience
-
Access to rider support
-
Privacy expectations
-
In-ride display experience and information granularity
The insights and voices that emerged from these sessions helped align cross-functional teams, foster shared understanding, and shape the development of Aurora’s rider-facing platform.
-
We interviewed seven self-driving car riders across San Francisco, Phoenix, and Austin to learn from their experiences with Waymo and Cruise and gather their stories. Among them was a rider who has taken over 800 rides with both services, and a rider who writes blogs and newsletters on Autonomous Vehicles.
.jpg)
Pre flight survey—the first exercise helped us build a foundation for 60 minute interviews by capturing participants' stories and overall impressions of todays self-driving car experiences.
Sacrificial concepts—the first exercise helped us build a foundation for 60 minute interviews by capturing participants' stories and overall impressions of todays self-driving car experiences.
Trade off exercise—we concluded with a trade-off exercise to better understand user priorities, preferences and values.
.jpg)
Insights and Learnings
Learning 01
Smart means understanding context and eliminating steps—Riders naturally compare the self-driving experience to familiar services like Uber or Lyft. Any extra steps required in an AV ride feel unintuitive. Riders expect the vehicle to simply understand and respond intelligently.
Thomas - First about unlocking the door, I kind of expected it to just be like … I mean, the car is so smart, right? It should just know that it's you and the door should just open. Matt : "As long as you tell me with the indicators, it's going left, right going right, and I can hear it and can see it, that would be nice."
Learning 02
There’s a key difference between providing information and delivering the right information at the right time—riders want relevant, timely information and expect to be accommodated based on contextual awareness; repetitive content adds no value.
Joel - Waymo hasn't figured that out. They say the same voice over lines over and over and over and over. And they have the same information on the screen. I don't know all of that. I mean, they have shortened it over the years, but it's still like, I just want to get in the car, leave me alone. Matt - I typically don't like to have it up with me and like it in a trunk, but sometimes maybe I want it with me and I think it'd just be nice when it's approaching and it says, would you like to open the trunk? Yes or no. Um, and it doesn't automatically open it because what if I'm in a rush and I just want to get out? I don't
Learning 03
Synchronize AV light patterns with universally recognized hazard signals to ensure clarity and reduce rider confusion—riders universally recognize blinking hazard lights as a signal that the vehicle is stationary and safe to approach. Introducing unfamiliar communication styles can cause confusion unless explicitly explained.
James - It is very clear to me that I can go or I need to wait. I'm not going to know what that AV flashing light means. Maybe it means the vehicle is not ready. Maybe it's doing the opposite job of what it's trying to do. When a hazard light is on, it typically means there's something wrong with the vehicle or it's paused. Joel - Yeah, the I think yeah, use of the hazard lights is probably pretty smart, I would imagine because that's universally understood as like, hold on, I'm, I'm gonna pause here for a minute for some reason.
Learning 04
Value of autonomous vehicles come from consistency and knowing what to expect at every ride—experienced riders emphasized the value of controlled and consistent experiences across every ride, from how the cabin feels and smells to how the vehicle behaves. This reduces unpleasant surprises and sets clear expectations for what to anticipate.
Joel - I didn't really appreciate the consistency of the experience until I started taking more Lyft and Uber. Then I was like, oh yeah, there's a huge value in knowing exactly what you're going to get before you get it. Maya - Every single time you get in, you know exactly how it's gonna drive. I just I know the way my driver very well, so I know what it's gonna do.
Learning 05
First and lasting impression of cleanliness comes from the way the cabin smells—the cabin should smell fresh and clean, with a pleasant and consistent scent, either distinctive or neutral. Fragrances that evoke hospitals or rental cars, can feel overly disinfected and impersonal.
Joel - One thing other subtle touch that Waymo used to do, whether on purpose or not, is the cleaner they used in their cars. They had a pleasant scent. So it's like, oh, that’s the Waymo scent. Thomas - I think it's the thing that sort of hits me first and the thing that's most memorable. But I feel like there's a lot of Ubers and Lyfts that I've taken recently that are like older cars … but yeah, usually it's the smell that gets me.
Learning 06
Riders rationally understand the need of a barrier but associates it with negative connotations—riders often associate barriers with police vehicles or pandemic-related precautions. It's important to design a barrier that feels open and approachable, while also allowing riders to clearly see the vehicle’s controls in operation.
Apolline - It's kind of funny, I associate the barrier with COVID because when all the Uber drivers were putting those plastic barriers in their cars to keep them safe. Joel - Nowadays, I’m not in favor of the barrier mostly because they are never clean and obscuring the vision out the front is fine…
Learning 07
Focus on maximizing space, not time—for riders, an elevated experience is tied to comfort, not productivity. Rides are often quiet moments to reflect, scroll, relax, and recharge before travel or work, not opportunities for calls or emails.
Thomas - The only work I will do on a flight is... I will just do expenses on my phone, if it involves anything with my laptop, I'm like not about it. Maya - One of the things I really like about sitting in front in the Waymo is that I can extend my legs and across my legs at the ankle which feels very comfortable.
Learning 08
Riders want to connect and play their personal media during rides. Wirelessly—there’s an unmet need for seamless wireless media connectivity in AV rides. Riders expressed frustration at not being able to play their own music or podcasts during private trips.
Apolline - When I first took the cruise, I thought you could play your own music. And I was really disappointed that it was just had preset radio. It just really silly to me that that wasn't a feature. It's so easy. Thomas - I like the idea of connecting your personal device and listening to your own stuff. Cause all they had in the Cruise was like the preset radios. And every preset radio I'm just listening to ads for all sorts of garbage that I didn't want.
Learning 09
Granular cabin controls would nice to have but just about right is enough—riders appreciated the ability to preset ambient lighting, music, volume, and climate settings. However, above all, maintaining a comfortable temperature was the most valued.
Matt - I would love temperature. Austin is so hot. So having it pre-cold would be nice. And I know with Cruise, some people left it really hot or really cold just to mess with the next person because it wasn't adjusted back down to like 70, Thomas - I don’t know if I’m too particular about the temperature or the light or the music. It would just be nice to have control over those things.
Seeing the way the steering wheel turn provides a sense of predictability—participants shared that seeing the steering wheel turn provides a sense of security and offers a visual cue for how the vehicle will behave. It adds clarity, enhances rider confidence, and even introduces a fun element to the experience.
Learning 10
Joel - I really like to be able to see the steering wheel move. I value seeing the twitch of like, oh, what are we going to do? Oh, we're going to go this, you know? Thomas -Having that visibility of the steering wheel gives you that feeling as if, you did have control. You would know how to control or you would be able to control.
Signaling upcoming actions before they occur helps riders anticipate movement and reduces unexpected surprises—first time riders inspect and evaluate. Riders want gentle cues before an actions occur. Turn signal sounds, path projections on the map, and contextual notifications serve as subtle signals that help reduce surprises and lead to building trust.
Learning 11
Matt - As long as you tell me, like with the indicators, it's going left, it's going right, and I can hear it and can see it. That would be nice. Because a Tesla with full self-driving, it will auto change the lane for me. And being able to know that it's doing it before it changes is helpful. Joel - It's absolutely huge for me seeing the rider-car communication. Having the screen that shows you exactly what the car is thinking, where it intends to go, where it knows it's safe to drive, where the next turn is, what the stop light says, all of that.
Riders trust that physical buttons are dependable and quickly accessible in emergency situations—though not frequently used, physical buttons were largely preferred over digital ones. Riders valued their tactile feedback, ease of location, and the reduced risk of accidental activation, contributing to a more confident and intuitive experience.
Learning 12
Joel - Physical buttons are nice. Trying to get a touch screen when [....] we are going around a corner or something. So physical buttons - props for those. Apolline - …If something really went down, I think I would rather press a real button for help and pull over in case of an emergency.
Riders take AV rides with the assumption that they will be monitored for safety purposes—interviewees responded that they enter the vehicles with the assumption that they will be monitored for good or safety purposes.
Learning 13
Maya - That's just how it is. I think that it's a safety thing. They say they also check the vehicle and the vehicle status when you exit. Some people believe that they're monitoring pickups and drop-offs all the time. I think there's a bit of unknown there and it didn't doesn't bother me. Matt - The ride is safer when there is the ability for someone to go back and see what happened in that ride. I am kind of in this like post 9-11 world, you know that you're going to be watched. So like, I already am accepting of that.
Questions and discomforts remain around how, when or what is being recorded—questions remain around the methods, purpose, and storage of data collection. While there is some skepticism about transparency, riders expressed appreciation for being informed when and how they are being recorded.
Learning 14
James - Most ideal is not being recorded, right? Second is, if you're going to record me, please let me know. If you don't let me know, it is going to affect my experience negatively. Thomas - Rationally, I knew that they weren't capturing my credit card number or something, but I was trying to order something for work and I was like about to put in the credit card number and I was like, this is kind of weird. I stopped and saved the cart and I put it away and I was like, I'll do this when I get home.
Presence of cameras and microphone should be communicated, not half hidden
—hidden cameras raise concerns about transparency and create distrust around how recorded data is used. Riders value clear, honest communication about the presence of cameras and microphones.
Learning 15
Joel - I like that the camera and microphone are clearly communicated that they're there and watching. Waymo used to be more clearly communicated but now the now the cameras are kind of stealth hidden. You can see them but they're they're not like promoted that they're there you know so.. Thomas - It was a little weird to me that they didn't like point out the cameras. In the Cruise car, they were trying to hide it, which is what kind of made it weird. If it's hidden and then you discover it, it kind of makes you start thinking about like what they could be using it for.
Reassurance comes from first acknowledging an issue, then providing a clear and thoughtful explanation—during unexpected movements or experiences, riders want to feel that the vehicle is aware of its behavior and appreciate an explanation to feel reassured.
Learning 16
Joel - There's been a couple of times on a Cruise where there's been random unexpected braking events. And it just doesn't say anything. It's like, nothing happened at all. Let's just continue. Like, wait, hold on a second. What was that for? Apolline - When the car does make some sort of unexpected movement, it should let you know what it's trying to do. I think that would make me feel more at ease.
There’s a need to clarify, educate, and set expectations around the terminology used in rider support—riders expressed uncertainty about the purpose and outcomes of the safety buttons, leading to confusion and anxiety.
Learning 17
Maya - The problem with pullover button is that it's not really clear to people. It really needs to be clear. It wasn't clear with Waymo. I don't use it. I think of it as like, oh my gosh, I left my keys on the roof of the car or whatever. That doesn't make sense, but something like that, something urgent. It feels urgent Matt - The only thing is, like, help, is it still driving when I request help? Or is it, am I requesting help because it's stuck? I only requested help when it was stuck or was paged in, it thought that the, like a pothole was something that swerved and it made me nervous, for example.
Opportunity Areas
Enhance rider trust by providing clear, contextual cues about vehicle behavior, enabling users to anticipate actions and feel more secure—first time riders closely evaluate the vehicle's behavior before letting their guard down. Building trust early is critical to helping them feel safe and at ease."
Thomas : "as you get going, you realize like, oh, it does know what it's doing. It was almost like you were seeing what the car was seeing. So you know by looking at the screen that it knew that things were there and that was kind of comforting.
Respect rider privacy by ensuring transparency and clearly communicating the status of monitoring systems—riders enter self-driving cars assuming they will be monitored for safety and development purposes—and generally accept it. However, they express a desire for transparency and clear notifications when monitoring is occurring.
James : "Most ideal is not being recorded, right? Second is, if you're going to record me, please let me know. If you don't let me know, for some reason, it is going to affect my experience negatively." Thomas : "It's a little weird that they didn't point out the cameras in the Cruise car. They were trying to hide it, which is what kind of made it weird. If it's hidden and then you discover it, it kind of makes you start thinking about what they could be using it for."
When navigating high-stress environments like airports, especially with bags in hand, riders want experiences that feel seamless and instinctive. Designing for this means leaning into familiar interaction patterns, offering flexible task flows, and providing clear next-step cues.
Joel - Waymo hasn't figured that out. They say the same voice over lines over and over and over and over. I mean, they have shortened it over the years, but it's still like, I just want to get in the car, leave me alone. Matt - I don't know, do I want to deal with the hassle of opening it or, umm.. it's like at a grocery store, you may have a preconceived idea of what you want to have for dinner, but then you change your mind when you're there.
Offer a spacious and personable ride focused on relaxation and quiet—Riders seek a simple, spacious environment where they can relax and recharge. This doesn't mean more features, but rather only the relevant information, easy access to personal music, a cool, clean cabin, and the space to be left alone.
Maya - "One of the things I really like about sitting in front in the Waymo is that I can extend and cross my legs to get comfortable." Apoline - "When I first took the Cruise, I thought I could play my own music. I was really disappointed that it was just preset radio. It felt really silly to me that it wasn't a feature. It's so silly." Thomas - "The only work I will do is basically expenses on my phone. If it involves anything with my laptop, I'm not about it."
Set clear expectations around the use and outcomes of support features to help riders feel in control—riders often lack awareness of available safety features designed to help them feel secure. There’s an opportunity to set clear expectations around their use and outcomes, so riders feel acknowledged and can access them appropriately with confidence.
Maya - "I think the problem with pullover button is that it's not clear to people. It really needs to be clear. It wasn't clear with Waymo. I don't use it. I think of it as like, "oh my gosh, I left my keys on the roof of the car or whatever." That doesn't make sense thought. Something urgent, that's the word, urgent. It feels urgent." Matt - "The ability to pullover different from requesting help would be nice. The only thing is like, help, is it still driving when I request help? Or is it, am I requesting help because it's stuck? Do I only request help when it's stuck or there was a pothole or something that swerved and it made me nervous."
Physical, digital, and service touchpoints come together to shape the rider experience—highlighting key moments to capture our key learnings
Respect the rider’s mental model—a self-driving car experience should feel no different from being picked up by an Uber or Lyft, only simpler and more seamless to start the trip.
Respect the rider’s mental model—a self-driving car experience should feel no different from being picked up by an Uber or Lyft, only simpler and more seamless to start the trip.
Consistency is highly valued—riders appreciate knowing what to expect on every trip: which car will arrive, having the same driver(driving behavior), and even the familiar scent of the cabin.
Focus on maximizing space, not time—riders aren’t aiming to be productive during the trip, but rather to enjoy peace and quiet.
Seeing is believing. When riders can see what the driver sees—and understand how the car is making decisions—they begin to trust it. Transparency in the car’s perception and behavior helps riders set expectations, evaluate its actions, and ultimately feel comfortable enough to relax.
Let riders play their media like music or podcast. This is not a nice-to-have, it's a requirement.
Respect rider privacy—when alone in the car, riders see it as their personal space. They understand cabin monitoring for safety but want clarity about when they’re being observed and what is visible.
Exterior design—transferring the sensor hardware from trucks to passenger vehicles
Interior design treatment—an approachable form paired with vibrant color contrasted from its surroundings to invite riders and clearly signal points of interaction.
This was truly a collective effort from a small team that moved quickly and decisively.
Insights from both self-initiated and external user research quickly aligned cross-functional teams and informed key decisions. The process was energizing, and we’re extremely proud of what we achieved in such a short time.
