GM Future Vehicles

In a sponsored project with Carnegie Mellon Design, General Motors asked us to research and design future interaction scenarios with semi-autonomous vehicles. I worked with Renee Yang to do a brief research phase, develop preliminary ideas, and refine a final concept.


View a demo on vimeo

Cars that drive themselves

General Motors worked with CMU's junior industrial design class over the course of a few months on research and interaction design concepts. They were interested in the unique interaction scenarios that could result from the commercial availability of semi-autonomous vehicles.

Research phase

We spent some time researching and creating models of the existing driving model, to both inform our concepts and give us a vocabulary for discussing them. We used a school survey, contextual inquiries, and make tools to diagram relevant parts of the experience of driving, which we used to direct our ideation phase. From our research we were able to draw conclusions about attention demands, social interactions inside of and between other cars, rituals and habitual behavior, and dangerous situations. We also took steps to familiarized ourselves with the state of autonomy, artificial intelligience, and display technology.

Preliminary design concepts

Communal space

The configuration of space and people in a car's interior is based on physical requirements that could change with semi-autonomous vehicles. The steering wheel/dashboard/windshield system could be replaced with a high level control device. The driver is no longer confined to the "driving position" and is free to engage with passengers in a communal space.

High level control

Computers could delegate direction to human drivers but manage lower-level driving tasks itself. This would relieve drivers of much of the cognitive burden of driving as well as allow for new input possibilities.

New control device

With the computer handling lower level driving tasks (ex. the minutiae of executing a left turn vs the idea of a left turn), the control device no longer needs a physical connection to the car's movement. At the same time, this creates an opportunity to unite the large orchestra of controls currently demanding the user's attention (gear shift, temperature control, A/V controls, etc.) and provide them with other functions, possibly now including entertainment.

Augmented driving

Our concept assumed a flexible degree of autonomy. Sometimes drivers would want to or have to take full control, while other times they would simply designate the destination. Advancements in display technology and computer vision could culminate in heads up displays overlayed on the car's viewport. This could make driving safer and easier even when the driver took full control.

Final concept

Our final concept pulled from our ideation phase to create a cohesive interaction scenario. A tablet would serve as both a high-level steering device when mounted in a pilot position, and detach to become a monitoring & control device so the driver could move about the vehicle.

The concept of autonomous vehicles hinges upon the idea that computers are—or will be—more competent than humans at safely driving cars. We wanted to bring this idea to manual driving. When making a left turn, there's the intent of turning left as well as the physical maneuvering of the vehicle within a safe path while avoiding other cars, an exercise in spatial awareness and timing that often causes accidents. Intelligent cars could dramatically reduce human error by making driving gestural—with gyroscope and accelerometer equipped tablets.

When autonomous mode is engaged, the tablet helps the driver monitor the car with a camera feed of the vehicle, controls and diagnostics, and provides entertainment via a basic web browser.