Screen Shot 2018-02-08 at 11.32.00 PM.png

Copy of Carpt


Carpt is a concept for an autonomous platform, designed for an emerging moment of transition between traditional transit and autonomous vehicle use. Autonomous vehicles are being touted as the future of transportation, but large-scale adoption of self-driving technology will not be seamless: from consumer behavior to civic infrastructure, interwoven systems shape the transportation landscape. Carpt can help facilitate flexible movement within this landscape. We conceptualized a fleet of on-demand autonomous platforms, on which traditional vehicles could be easily mounted and transported.

User Research

We embarked on wide-ranging user research, conducting observations and interviews. We interviewed a cross-section of people, from an undergraduate student to a local bus driver. Methods like empathy mapping and concept clustering helped us see common themes across this research. One theme that emerged was people’s discussions of the need for comfort, particularly in relation to elements of transportation that felt uncertain or out of one’s control — whether that was a discomfort with strangers on the bus, uncertainty about when an Uber driver would arrive, or anxiety over unexpected traffic. At the same time, we noted a need for a diverse ecosystem of transit modes and solutions, as people’s individual routines and needs varied widely. Together, these insights led us to ask a question that would frame our next steps: “How might we design transportation systems that transform uncertainty into excitement?”

IMG_2235 (1).JPG

User Interviews

IMG_2247 (1).JPG

Empathy Map

IMG_2270 (1).JPG

Point of View Statements

How Might We Statements


How might we design mobility systems that transform uncertainty into excitement?

These insights led us to brainstorm a range of early concepts, centered on building an understanding of comfort and excitement into transit systems. As we imagined these concepts, we were drawn to designing for an emerging moment, one in which autonomous vehicles will become increasingly prevalent. The transition period to a population of purely autonomous vehicles will create a complex ecosystem in which human-driven and fully autonomous vehicles share space, presenting challenges for drivers, vehicle companies, and policymakers. We thought about how we might ease this transition while accommodating the needs of a diverse user landscape, and where there might be an opportunity to create comfort and excitement within this period of uncertainty.

Final Presentation.jpg

Some Ideas

Transformable Clothing for Bikers

Transformable Clothing for Bikers

Autonomous platforms for ordinary cars in cities

Autonomous platforms for ordinary cars in cities

Autonomous Cars do not stop at intersections

Autonomous Cars do not stop at intersections

Themed cars for public transit

Themed cars for public transit

Concept Generation

Final Presentation (1).jpg

Our concept positions Carpt in the relatively near future, in a period when autonomous driving technology is available and safe; some people own autonomous vehicles, but most don’t; and some cities or areas have restricted vehicle use to self-driving cars. This transition period will create a complex ecosystem in which human-driven and fully autonomous vehicles share space, presenting challenges for drivers, vehicle companies, and policymakers. Considering this ecosystem, we thought about how we might be able to ease the experience for stakeholders — and where there might be an opportunity to create comfort and excitement, two key values gleaned from our user research, within a period of uncertainty.


Use Cases

Final Presentation Skeleton - Ford.jpg

In the personal transportation space, we imagine diverse use cases, including driving conventional cars in areas where human-driven cars are eventually prohibited, such as cities that may adopt new policies in the coming years; use for long-distance travel, when a driver may want to rest or work; and use for scenarios in which a trailer might currently be used (for example, transporting a vintage vehicle to a car show without adding mileage).

Final Presentation Skeleton - Ford (1).jpg

More broadly, we envision Carpt becoming the center of an ecosystem of services. A “CarGo” service, for example, could allow customers to order a Carpt platform with a cargo container attached — one could easily imagine both individual uses of this, like moving, as well as commercial uses in ground shipping. Along similar lines, a “CarPod” service that packaged platforms and cars could be used as an autonomous cab or ridesharing service. Carpt would serve as the shared element across these modular, on-demand offerings, helping smooth the transition to a new mobility era for individuals, businesses, and governments.


We began prototyping the Carpt hardware by brainstorming and sketching potential system designs. We discussed features we thought would be key priorities, using sketches to visualize and iterate on our ideas.


Conceptual Lego Model

We started with a small Lego platform and toy cars, a simple way of tangibly demonstrating the basic concept without diving into details. This allowed us to illustrate our idea to potential users early on in the process, framing open discussions about potential use cases and other feedback.

Foamcore 1

we began to build more detailed prototypes using foamcore. These demonstrated actual mechanisms — such as two-degree-of-freedom suspension arms, the loading process, and the battery pack — and the platform’s form factor.

Foamcore 2

These prototypes were key as we conducted more user interviews; they helped us gather feedback on aspects like safety concerns, the volume footprint, and the design’s complexity. We incorporated this feedback in our second iteration.

CAD Models

We moved into higher-fidelity CAD modeling and animations, refining the Carpt system and its features.

Hardware Features

  • Wheels are in front of and behind the platform (rather than on the sides), allowing it to make maximal use of the allowed width for road vehicles.

  • A symmetric design allows the platform to move in either direction, simplifying boarding and unboarding.

  • The suspension design allows for all-wheel steering; lowering the platform (for loading/unloading) or lifting it up (for 90-degree parking); leaning into curves; and actively stabilizing the platform in a z direction.

  • The system uses an all-electric powertrain, with four independent hub motors.

  • Latches automatically attach a car on the platform on the inside of the wheels, and automatically adapt in length and width to fit most personal vehicles.

  • An adapter unit allows for 12V charging and connecting the built-in AC to the passenger’s car ventilation.

Mobile Application

Paper Prototype

We prototyped the app that would allow users to call and manage the platform. We started by hand-sketching screens on sticky notes, mapping out potential user interactions. We used card sorting method to streamline the application screens.

Sketch prototype 1

As we moved forward, we incorporated our hand-drawn sketches into Sketch to create more realistic renderings of the app’s interface.

Sketch prototype 2

Throughout this process, user interviews and testing helped us iterate on the app’s features — we added functionality around where the Carpt would wait for a user, how the app would help guide a user in docking their car, and how the user could customize their Carpt to display a message that would uniquely identify it when it picked up a user.


Interaction Design

We then used Flinto to bring the app to life, allowing for touch-based interactions on mobile devices.

App Features

Users can...

  • Specify the summoning and destination points

  • Specify a location for Carpt to wait for them (such as instructing it to stay at a spot or to drive around while they finish something up)

  • Follow intuitive visual guides to dock their car

  • Pre-load cars into their account, so that Carpt knows the type of car it is transporting

  • Order Carpt for cargo shipping or vehicle loading

  • Customize the message that Carpt will display to notify them that it is their Carpt

  • Tell Carpt to auto-pak

1 (1).jpg

User Testing

The cyclical process of iterating and gathering user feedback played a crucial role in helping us define Carpt’s core features. From our testing sessions and conversations, we heard several key themes. Some user feedback focused on how individuals would use and experience Carpt: people mentioned that they might be more likely to use Carpt for long drives than for a quick 30-minute errand, for example, and we heard questions about on-the-ground concerns like passenger safety, maneuvering narrow roads, and storage of the platform. Feedback pointed to questions suggesting that we think about how an on-demand platform service would be different from autonomous ridesharing and how it might be applied to business or infrastructural processes, like transporting cargo.

Final Presentation (2).jpg

A key moment came when we found a solution that would both address many of the concerns we heard from users (like loading the car, maneuvering or parking in tight spots, and attaching the car) and significantly simplify our design. At the crux of this solution was the idea to locate the wheels in front of and behind the platform, rather than on the sides. This streamlined the platform’s features, allowing us to move toward the design that ultimately became Carpt.


Featured Links:

Carpt was created as part of Reimagining Mobility, a course offered by the Jacobs Institute for Design Innovation at UC Berkeley and the Ford Research and Innovation Centre.

Team: Advaita Patel, Anosh Sethna, Lexian Guo, Patrick Hartmann