A recent Marketplace article, “Your Car Is Not Self-Driving, No Matter How Much It Seems Like It Is,” quotes informatics Ph.D. student Hillary Abraham in response to concerns from safety advocates that driver-support technologies will be viewed as self-driving. “If that’s the attitude, then it’s only a matter of time until you get in an accident,” she explains. The article reports on a study Abraham conducted while working as a research assistant at MIT’s AgeLab. She found that car dealerships don’t always sufficiently explain to customers how advanced safety features work. Now at UCI, Abraham is exploring broader questions about the future of transportation from a variety of perspectives.
Can you talk about your path to UCI and decision to get a Ph.D. in informatics?
I was working on a project at MIT investigating how drivers interact with existing technologies. The project focused on what drivers did, but I became very interested in why drivers did what they did, and what that meant for the future of transportation. I decided to pursue a Ph.D. that would allow me to explore these questions. UCI was a unique choice for me. We don’t have a lab here focused on driving. However, the Informatics Department has exceptional faculty in social science, computer science, design, anthropology, education, sustainability and policy, which all intersect the broad topic of the “future of transportation.” Here, I have the opportunity — and am encouraged — to explore the facets outside of the car that impact what happens in the car.
Why are you interested in this area of research, and what is your current focus?
We’re in the midst of a potentially large shift in how we move around, and times of change are always interesting. Additionally, driver assistance and automated vehicles lie at the intersection of many different fields. This gives me the opportunity to learn about a variety of topics and work with a variety of individuals. Recently, I’ve been exploring how automation in domestic contexts has changed relevant tasks and lifestyles. I’m hoping this will help add context to how automation is applied in vehicles.
What are your biggest concerns around driver assistance systems and automated vehicles?
Some of my research has identified that people are really confused about driver assistance systems and automated vehicles. When purchasing vehicles, it’s unclear what systems are available, what systems are actually in the vehicle, what their purposes are, how and when to use them, and their limitations. People often end up turning them off or using them improperly. These technologies have great potential to save lives, but only if they are used — and used appropriately.
Are there ways to better design the technology, or do we need to better educate users?
The solution probably lies somewhere in the middle. We’re already really good at designing technology that people can learn to use very quickly — smartphones are a great example. You don’t need much (if any) training to figure out how to do basic tasks. You’re able to make mistakes and recover quickly, and consistent feedback helps improve your understanding over time.
There are a few differences with vehicles. First, you don’t have an opportunity to make mistakes safely. When you’re driving, you need and want to know how the vehicle is going to react before you try anything. In this case, education can help users understand the system before they actually use it. Second, your vehicle’s response can depend on the situation. For example, a car might respond different in rain versus in clear weather, but it might not be apparent that the weather is causing the change in response. This might be difficult to teach, because you can’t expect users to remember all of the situations in which a system might not work properly. This is where an improved interface design can help.
This all falls under the larger umbrella of human-computer interaction. What is the biggest challenge in HCI as we face increased automation?
I think calibrating the level of automation, or deciding when to let the human make decisions, will be challenging. My research in vehicles has demonstrated that people vary widely in what level of control they want to have over the driving task, and I imagine this will be similar for different forms of automation as well.
What are your future plans?
At this point, I could see myself in either an academic or industry position. Because automation is such a broad topic, I can explore it through a variety of lenses. My goal right now is to get exposure to those different lenses, which will help me decide on my dissertation topic and inform my long-term career goals. What’s great about being in this field at this time, and studying at UCI especially, is that I have the option to explore these multiple different career paths. I’m excited to see which path I take!
— Shani Murray