Don’t delay: Early engineering intro pays off

When Mohit Bhardwaj was a freshman in high school he trav­eled from his home in Lusaka, Zambia to Boston. with Lead America. For nine days, he and a group of 19 other stu­dents from around the globe studied lead­er­ship and engi­neering at Olin col­lege. They learned to work on teams building Lego robots, pop­sicle stick bridges, and the classic egg drop device. Bhardwaj’s team designed a con­trap­tion that allowed their egg to spin gently to the ground without breaking. Four years later, he found him­self back in Boston, a full-​​time stu­dent at North­eastern, using the same tac­tics to suc­cess­fully guide a pumpkin from the roof of Columbus parking garage to the ground below.

This summer, after a quick visit with his family back home, Bhardwaj has been working with a couple of cur­rent high school stu­dents in the lab of mechan­ical and indus­trial engi­neering asso­ciate pro­fessor Rifat Sipahi. The stu­dents, Pel­le­grino Conte and Maria­mawit Loulseged, were par­tic­i­pating in the Young Scholars Pro­gram through the Center for STEM edu­ca­tion. Nei­ther of them knew how to code when they got here, but they quickly learned with the help of Bhardwaj and two grad­uate stu­dents, Payam Parsinejad and Payam Nia, from Sipahi’s lab.

This was for­tu­nate because the stu­dents were tasked with designing a video game that would help Sipahi and his team under­stand how humans interact with delayed sys­tems. We’ve all expe­ri­enced it — you’re watching a “live stream” online of some­thing that hap­pened in real life five sec­onds ago. Your cell phone con­nec­tion is a little wonky and it takes your friend a few extra moments to respond, irra­tionally trig­gering  your age-​​old fear of rejec­tion as you wonder if your friend has just sud­denly hung up on you.

SAM_9216
Pres­i­dent Joseph E. Aoun and Mohit Bhardwaj at an event to thank ori­en­ta­tion leaders last week. Photo cour­tesy of Mohit Bhardwaj.

These are the kinds of things Sipahi’s lab is inter­ested in. They want to under­stand these kinds of sys­tems with delays–which man­i­fest in every­thing from pro­tein syn­thesis to oper­ating the Mars rover from Earth–in order to better con­trol them. Just as a bad cell con­nec­tion could be the first step in a tum­bling cas­cade of com­mu­ni­ca­tion break­down, trig­gered by your need to just feel wanted and accepted, delay can have some pretty bad con­se­quences in other types of sys­tems too–imagine how bad it would be if there were a delay in Bhardwaj’s pumpkin drop project. “You can never elim­i­nate the delay,” Bhardwaj told me, “but you can try and make proper actions to keep things in control.”

The com­puter game he helped Conte and Loulseged develop is simple: a square moves on a random and con­tin­uous path around the screen and the player tries to “catch” it with a circle whose path he can con­trol with a haptic device. The tricky part is that the circle can be a little lazy. In one iter­a­tion of the game, it takes it a half second for the circle to carry out any com­mand the player gives it.

To over­come the serious dif­fi­culty of chasing some­thing with a blind­fold on, which is essen­tially what this game feels like, the pro­gram­mers added in a “spring factor.” The center of the circle and the center of the square are attached by an invis­ible string (just like Jane Eyre and her beloved Mr. Rochester) that con­tin­u­ally draws them back toward each other. The far­ther you pull the circle away, the more it wants to snap back to the square. The devel­opers made it so the player can chose what ingre­di­ents she wants to play with: spring or no spring, delay or no delay, etc.

While they haven’t ini­ti­ated a con­trolled study yet (they only just fin­ished the game a few days ago), the team has already noticed, per­haps unsur­pris­ingly, that playing the game with a delay and no spring factor is ridicu­lously frustrating, whereas with the added spring factor, the user does not get frus­trated as much and can com­fort­ably meet the game objec­tives, sug­gesting that the invis­ible spring smoothly assists the user, said Sipahi.

Sipahi’s lab is working to build var­ious tech­niques into delay sys­tems to keep them from going totally hay­wire. But when these sys­tems involve some ele­ment of human inter­ac­tion, the issues of psy­chology and behavior come into play. With games like this one, the team can study these ele­ments and begin wiring them into their systems.

 

Related Faculty: Rifat Sipahi

Related Departments:Mechanical & Industrial Engineering