There’s no right pace,” said Mohamed Kante, E’12, who worked with elderly and dis­abled patients at Kin­dred Tran­si­tional Care and Reha­bil­i­ta­tion — Craw­ford in Fall River, Mass. No matter how fast or slow he and his col­leagues offered patients bites of food, they could never match the patients’ indi­vidual needs.

So Kante and five of his elec­trical and com­puter engi­neering class­mates decided to solve that problem with a senior cap­stone project that puts the con­trol in the patient’s hands, or — in this case — their eyes.

The under­grad­uate student-​​researchers won this year’s first-​​place award in the ECE cap­stone com­pe­ti­tion for devel­oping an eye-​​controlled robotic arm that allows patients to feed them­selves. “Once they have the ability to do it them­selves, there’s an enor­mous sense of freedom,” said James Barron, who devel­oped soft­ware for the project.

The cap­stone team included Nick Aquino, Barron, Kante, Ryan LaVoie, Pedro Lopes and Basel Mag­fory. Waleed Meleis, an asso­ciate pro­fessor in the Depart­ment of Elec­trical and Com­puter Engi­neering, served as the group’s fac­ulty advisor.

The eye Con­trolled Robotic Arm Feeding Tech­nology, or iCRAFT, has the poten­tial “to give thou­sands of par­a­lyzed indi­vid­uals the inde­pen­dence to eat with min­imal help from a care­giver,” Meleis said.

Sim­ilar tech­nolo­gies exist, including the recently reported Brain­Gate implant, which allows patients to con­trol a robotic arm merely by thinking about it. But these  require some kind of inva­sive — or even sur­gical — inter­face to con­nect the user’s desires with the robot’s behav­iors, Lopes said.

In this case, there is no phys­ical con­nec­tion between the user and the con­trol device — no joy­stick under their chin, for example. Instead, the patient needs only to look at a box on a com­puter screen.

The team devel­oped an eye-​​tracking soft­ware that cou­ples the direc­tion of a patient’s pupils with his or her food choices. Three col­ored seg­ments of the screen cor­re­spond to two bowls of food and a drinking bottle. A fourth, larger seg­ment allows the patient to take a break from eating.

Meleis said that the graph­ical user inter­face designed by the team is impres­sive because of its sim­plicity. The judges, 12 prac­ticing alumni engi­neers, “were par­tic­u­larly impressed with the impact iCRAFT will have on the target pop­u­la­tions and by the suc­cessful inte­gra­tion of eye tracking, robotics, a custom GUI and spe­cial­ized equip­ment,” he said.

[media-​​credit id=9 align=“alignright” width=“350”][/​media-​​credit]

The single best moment of this cap­stone expe­ri­ence was the first time we were actu­ally able to con­trol the robot arm with nothing but our eyes,” Barron noted. “Once we were able to accom­plish this feat, I was con­fi­dent that every­thing else would fall into place.”

He was right. Aside from win­ning first place in the cap­stone com­pe­ti­tion, the team has devel­oped a tool that com­mu­nity mem­bers can use imme­di­ately with the appro­priate tech­nical know-​​how: The iCRAFT team has pub­lished the robotics plans online and the soft­ware package is avail­able as an open-​​source download.

Cur­rent alter­na­tive self-​​feeding devices cost in the range of $3,500, but iCRAFT can be con­structed for around $900, making the tech­nology a more afford­able option for dis­abled indi­vid­uals and their care­givers and families.