The future of autonomous commuting
PROJECT: Avivi is an autonomous shuttle service that offers commuters a fixed route in the morning and a flexible on-demand service after work. Avivi is designed to bring autonomous vehicles to the public in a way that is sustainable, affordable, and dependable.
CHALLENGE: Design an ecosystem that spans across multiple channels; 7 weeks
TEAM: Carlie Guilfoile, Ulu Mills
TOOLS: Adobe AfterEffects, InDesign, Illustrator, Fusion 360, Pen + Paper
OUR BIG QUESTION
How might we leverage autonomous vehicle technology to bridge the transportation gap sustainably, affordably, and dependably?
1 By 2030, 20M+ AVs are expected to be in operation. While this is reality, there are many possible futures that dictate how these vehicles will shape our cities, towns, and roadways.
3 Public transportation and city budgets have a limited reach. The last mile problem hinders people from taking buses, trains, and subways.
2 90% of U.S. commuters use personal vehicles as their primary mode of transportation – causing major road congestion during commute times.
Ulu and I partnered with a mutual interest in creating something for social good. During our early meetings, we discussed what current disruptive technologies have potential to have the most positive effects on communities. I shared my personal interest in transportation and energy and proposed that we think about how autonomous vehicles might shape our communities and roadways for the better.
We developed an Autonomous Vehicle territory map to discover shared concerns among stakeholders in the commuting space. After our first few iterations, we realized that visualizing these shared concerns was difficult because there was a lot of overlap between different assemblages of stakeholders.
Our final territory map reflected the fact that these lines of concern were blurred, by showing a group of users in the center, surrounded by their granular concerns, with overarching themes in the last layer.
SURVEYS: COMMUTING & AVS
We followed up with two surveys, asking commuters about their various attitudes and experiences with commuting and AVs. With 70+ respondents, we learned that:
Many people like public transit because it's a time of solitude where they can enjoy a book or podcast uninterrupted by the stresses of driving.
They dislike it because of its unpredictability, vulnerability to delays, and pains of connecting from one's home.
“I’m happiest when I don’t have to think about my commute and it just goes well.” –Respondent
We conducted 4 interviews with professionals at the forefront of autonomous vehicle research. While many couldn’t discuss their research explicitly, they helped us to begin asking the questions that are driving much of the AV research happening today.
RIDE SHARING COMPETITIVE ANALYSIS
We researched 5 ride sharing & transit services to determine our service model. We also did research on pilot AV services, like the Mcity Driverless Shuttle and May Mobility AV Shuttles in Detroit. We focused on identifying:
Much like Chariot, we wanted to have different types of payment options (All Access vs Credits) to make our service flexible, yet efficient. We also found interest in how Via licenses their technology to local transportation authorities. In combination with our scenarios and exploratory research, competitive analysis helped us lay a foundation for how we wanted to serve our user.
This wealth of information from stakeholders and experts, along with secondary research, helped us to formulate the design imperatives that guided us through the development of our service.
Our service must prioritize safety.
Our service should consider reliability and comfort.
Our service could address the bigger picture.
We developed four unique personas and scenarios that helped us think about what an autonomous ride-sharing service might look like for people of different income levels, different living situations and commuting needs. Another thing we were able to do here was begin to experiment with voice, visual and tactile features – and how they would fit into a system that spanned home, vehicle and work. Going into testing, we decided to try a branching storyboard model to gut check different components.
We tested with prospective users: commuters who own a car, but do not use it to commute to work. This helped us to discover which touchpoints would be most useful to our target audience. We learned that is that users wanted to be able to go on autopilot (a best case scenario when riding public transit currently) and be able to enjoy solitude (the greatest affordance of commuting by car), and just about everything else was superfluous. Anything we created would have to facilitate those things in order to resonate with our target users. Relating back to our imperatives, this translated to a slight reprioritization:
Comfort. On short commutes, people value “me time” over added features.
Reliability. To users, reliability means being able to get to work everyday at a certain time, but having flexibility on the way home.
Safety. In the context of a service, people are less concerned about safety than they are about ease.
This led us to build a multiple-touchpoint system that leveraged a home voice assistant for route updates, a mobile app to track, set and schedule routes, and a key fob for seamless entry onto the Avivi shuttle and other public transit systems.
The primary point of communication with the service is the app. Here, users can select their plans, plan and track routes, and receive notifications for most convenient connections.
To allow for personalization and safety, the Avivi key acts as a checkin device, and used in-shuttle can set personalization settings.
WHAT WAS HELPFUL
This is a vast research landscape at the moment, and it was a lot to take on for a six-week project. We were lucky to conduct it at Carnegie Mellon and in Pittsburgh, because it afforded access to conversations with people who could help frame the potential of AVs in a way that sifting through research papers alone could not. Our solution at its core is quite simple, but it only works with autonomous technology, and their insight was essential for framing it properly.
WHAT HAS THE MOST POTENTIAL
Based on research and testing, we believe that the service model, with a core customer base using the service in a scheduled manner supplemented by users who fill empty seats at the last minute, is a very economically sound use of AVs and can be of help to a wide range of users. This essentially is the model that many existing train services use, with both reserved and unreserved seats, and we see the potential for applying a similar model to help sustain a commuting transit service that gets regular scheduled usage.
GIVEN MORE TIME
We would have further developed our service-scape in three key areas. First, we would have developed an in-car screen component to communicate safety measures to the riders. We didn’t develop this piece for the presentation because we received mixed feedback around this component during user testing. We also felt like there was more research to be done around when to provide visual feedback to the riders. Second, given more time, we would have done more testing and prototyping on the SOS feature of the Avivi key. We explored this feature as a tool for interpersonal safety but ultimately didn’t feel like it was an essential element of our user’s system. Third, we would like to test more users in order to consider features and comfort levels for different lengths of commute.