Dinda and Memik Quoted in McCormick Article Discussing New Mobile App Project



Prof. Peter Dinda and Prof. Gokhan Memik have been quoted in a recent article discussing their mobile technology project, funded by the National Science Foundation and Intel, which involves creating an emotion-sensing interface that allows phones to gauge user satisfaction.

Excerted from an Thursday, September 11, 2014 article published by McCormick, titled, "New Phone Technology Could Sense Dissatisfaction."

You need to download an important email, but your smart phone is moving frustratingly slow. Your palms grow sweaty, and your pupils dilate in rage.

A McCormick School of Engineering team is working to develop phones that will understand these biological signals and adjust their performance accordingly. Something as simple as an accelerated heartbeat could let your phone know it needs to speed up.

“Your phone is continually making a trade-off between how fast it goes and how long the battery life should last,” said Prof. Dinda, professor of electrical engineering and computer science. “Right now, it makes those decisions with no input from you whatsoever.”

Prof. Memik, associate professor of electrical engineering and computer science, and Prof. Dinda are creating an emotion-sensing interface that allows phones to gauge user satisfaction.

“Ideally, you wouldn’t notice that the interface is even there,” Prof. Dinda said. “One way of doing that is emotion sensing. If you are irritated, then perhaps it’s because your phone is too slow.”

Funded by the National Science Foundation and Intel, Prof. Dinda and Prof. Memik built a patented system with biometric input devices, which monitor physiological traits, such as eye movements, body temperature, and heart rate. They hooked up study participants to the system to record physiological reactions to factors like phone speed and screen brightness. After a physiological trait changed, the team connected those changes with their causes, which could be the device’s performance or a reaction to screen content.

“If you know that the content hasn’t changed, then our system can tell that the change is because of device performance and not because of whatever you were reading on the screen,” Prof. Memik said.

While it might seem as though phones should always run as fast as possible, faster is not always better. Prof. Dinda and Prof. Memik assume that most phones perform higher than the average user needs because they are typically designed with only one type of user mind. This leads to short battery life, high energy consumption, and possibly a device that is unpleasantly hot to the touch.

“When you only try to satisfy that one type of user, you make other users much more dissatisfied,” Prof. Memik said. “Running at a higher performance than necessary makes your device more abrupt, and it becomes too hot.”

“One of the purposes of emotion sensing is we can tailor your phone specifically for you,” Prof. Dinda added.

Eventually, Prof. Dinda and Prof. Memik would like their hardware built directly into smart phones by developers. It could take advantage of information from health monitoring devices that people already wear or incorporate new pieces into the phone’s hardware, such as shake detectors and infrared cameras for recording pupil movement and dilation.

“Smart phone companies are already coming up with so many attachments,” Prof. Memik said. “There are smart wristwatches and wristbands. We can attach sensors on something people are already wearing. Our goal is to make it seamless, so they don’t even notice.”