As our pals at Grub Street reported earlier this week, Pizza Hut in the U.K. is experimenting with technology designed to ease the difficult Sausage or pepperoni? and Mushrooms or olives? decision-making process. The chain’s “subconscious menu” uses eye-tracking tech to predict with surprising accuracy what diners want on their pies — the system is right 98 percent of the time, the pizza chain claims.
The cost of eye-trackers has fallen in recent years, and, as a result, research using the technology has quickly grown. The scientists I contacted who use eye-tracking in their research on human behavior very politely hinted that they doubt Pizza Hut’s use of the technology will be as beneficial to society as some of the other research going on. And looking at what researchers are doing in this field, it’s hard to disagree:
- People who are paralyzed or have some degree of motor disability can use eye trackers to control a computer or operate a wheelchair.
- Medical researchers are developing ways of using eye-trackers to diagnose neurological conditions like ADHD, schizophrenia, and autism. There is some evidence that babies who are just six months old start to show differences in attention that could indicate autism, which can be detected with eye-tracking sensors.
- Surgeons may be able to use eye-trackers to operate robotic equipment in minimally invasive surgery.
- Sensors tracking eye movement can alert drowsy truck drivers, activating alarms and seat vibrators to wake the sleepy trucker back up.
Adrian Staub, a cognitive psychologist and director of the eye-tracking lab at the University of Massachusetts, Amherst, explained in an email how the sensors work:
If you are looking straight at me, your pupil will look perfectly round from my perspective. As your eye moves to the left or to the right, or up or down, it will appear to me that the shape of your pupil is changing, becoming an ellipse rather than a perfect circle. The eyetracker uses mathematical equations to figure out where you’re looking based on the exact shape of this ellipse.
Some eyetrackers can do this operation as fast as every millisecond, that is, 1000 times per second. This sounds like it might be a pretty rough calculation, but actually it’s very precise. The best eyetrackers can identify which letter you’re looking at as you read normal-sized text on a computer screen.
On a side note, it’s kind of amazing that the same technology that can aid in medical diagnoses can also detect the inner desires of British pizza eaters.