Human-​​machine interactions, the natural way

When mechan­ical and indus­trial engi­neering pro­fessor Yingzi Lin was pur­suing her PhD in vehicle engi­neering, one of the driver test sub­jects became so dis­tracted by the sensing equip­ment in the vehicle (the researchers were col­lecting data about his phys­i­o­log­ical state while dri­ving), that he got into an acci­dent. This expe­ri­ence got Lin thinking, “how could these exper­i­ments be as least dis­rup­tive for the test driver as pos­sible?” She began to explore that ques­tion in the con­text of research trials, but soon found that the appli­ca­tions were bound­less and applied to real-​​world sce­narios as well.

After earning her vehicle engi­neering degree, Lin moved to Canada to pursue yet another PhD in mechan­ical engi­neering at the Uni­ver­sity of Saskatchewan. In 2005 she joined North­eastern and the rest is his­tory. Well…actually, that’s not true. The rest is a story still being told. Lin cur­rently has about seven grad­uate stu­dents each pur­suing his or her own research in human-​​machine inter­ac­tions, and each with the ulti­mate goal of making those inter­ac­tions as seam­less and nat­ural as possible.

At the recent NSF Engi­neering Research and Inno­va­tion Con­fer­ence, hosted by North­eastern, Lin and her stu­dents pre­sented thin films that incor­po­rate a variety of phys­i­o­log­ical sen­sors into the mate­rials itself. Cur­rently the mate­rial is capable of detecting pres­sure and tem­per­a­ture inputs and now the team is working on get­ting them to detect blood volume pulse/​heart rate sig­nals as well. Lin orig­i­nally thought of the mate­rial as a skin to go over the steering wheel of a test car, so that researchers could col­lect data on the phys­i­o­log­ical state of a driver without inter­fering with his or her nat­ural expe­ri­ence of the envi­ron­ment. An added ben­efit, she said, is that the films do a better job of inte­grating the var­ious signals.

Lin is inter­ested in dri­vers because of the broad social impacts such an area has, she said. For example, vehicle acci­dents are still a leading cause of death despite advanced safety tech­nolo­gies and even infra­struc­ture devel­op­ment. If someone is tired or angry, cur­rent tech­nology can’t keep them from veering off the road or making a rash deci­sion in the heat of the moment. “About 16 mil­lion Amer­i­cans have road rage issues,” said Lin, who doesn’t expect that Amer­i­cans are alone here. If we can detect a driver’s emo­tional state, for example through ele­vated blood pres­sure and body tem­per­a­ture, we could poten­tially inter­vene at a crit­ical moment to either calm them down, wake them up or alert other dri­vers or offi­cials of a problem.

Of course, the car is not the only place where humans interact with machines. In our increas­ingly tech-​​savvy world, rarely an hour goes by that we don’t interact with a machine of some kind. Lin is pur­suing nat­ural sensing design in research projects as far reaching as phys­ical therapy inter­ven­tions, web site design and telemed­i­cine, to name a few.