The Dangers of Voice Controls

Guest Bloggers

Guest Bloggers | Oct 14, 2014

Voice-activated systems are getting more attention as states increasingly ban the use of handheld devices by drivers. But new research from AAA finds that these voice-activated systems can still be quite dangerous. National distracted driving experts Paul Atchley, of the University of Kansas and David Strayer, of the University of Utah-- and the lead researcher on the AAA study-- explain why regulations that favor voice-activated systems aren't making our roads much safer.

Car Talk: Where do you think politicians are getting it wrong?

Atchley: The biggest myth is that as long as our eyes are open and looking at the road, we are safe. 

It is not about eyes, it is about seeing and that requires your brain.  A person only needs to read a few crash reports where drivers say, "I was looking but I never saw the car/pedestrian/cyclist" to know the disastrous consequences of believing that looking and seeing are one in the same.

The lights are on, but nobody's home. (AAA)

Car Talk: Why is the myth that voice-recognition systems are safe so persistent?

Atchley: We underestimate how difficult talking is, because we have been doing it for so long.  But it is the most difficult thing your brain does.  It takes years to learn and it is the first thing to go if our brain is damaged in some way.

When we talk, our brain must take resources from other tasks like seeing. Just think about what someone sounds like when they talk on a phone and drive ("ummm… uhhh" and short, clipped responses). You are literally hearing the brain trying to take resources away from the conversation so it can see the road.

Car Talk: Why is Toyota¹s Entune system less distracting than other voice-recognition systems?

Strayer: We tested two “core functions” in our lab at the University of Utah: Voice dialing and radio tuning, because these are standard features on almost all new vehicles, and they’re related to the NHTSA manual radio tuning/phone dialing tests— tests that were primarily focused on visual distraction.  Some manufacturers have additional features (for example, climate control) and vehicles equipped with Siri or Apple’s CarPlay support yet more features, such as sending and receiving text messages.

Toyota's Entune system was less distracting than other systems, and had a workload rating similar to listening to a book on tape. It was less distracting, because the system was robust.  With Entune, there were few errors, the interactions were brief, and there was more than one way to get the system to do what the driver wanted.  Other systems had more errors, the interactions were longer, and drivers had to say the command in just the right way. (And even then, the system often failed. In one example, a driver attempted to change the radio, and the car changed the heating system!)

AAA’s takeaway is that the voice-based interactions should support the task of driving. Just because the technology is in the vehicle doesn’t mean that it is necessarily a good idea to use it when the vehicle is in motion. Social media, for example, is best left to non-driving situations. Deborah Hersman, the President of the National Safety Council put it well, I think when she said, "Infotainment systems are unregulated. It is like the Wild West, where the most critical safety feature in the vehicle - the driver - is being treated like a guinea pig in human trials with new technologies.”

(AAA Photo)

Car Talk: So, a return to the old days when drivers could only socialize with the people actually present in your car. What¹s the next study area in which distraction researchers will be focusing?

Atchley: To me the most interesting question is why a handful of studies that use the "naturalistic" method, using cameras in cars driven by volunteer drivers claim talking on a phone makes you a safer driver. This contrasts with over 400 studies over the last forty years using a variety of other methods, such as simulators, behavioral techniques, and observational/on road studies, along with neuroscience studies, all find that talking poses an increased risk.

The question is critical because industry has pointed to data from this one method to support an explosion of distraction in cars.


Get the Car Talk Newsletter

Got a question about your car?

Ask Someone Who Owns One