Mechanical engineering students using their heads to build brain-computer interfaces

Sitting in his apartment, Purdue University Fort Wayne graduate student Abdelrahman Gad is recording a video that looks like action from an X-Men comic book or a big-budget movie mimicking telekinesis.

But Gad’s efforts are not special effects, and his results are reality.

At his mental commands, a small, four-propeller drone lifts about three feet off the carpet and executes in-air turns around a box in the middle of the room. Gad is wearing a headset connecting to 16 contact points around his head, and the device uses a computer to translate his thoughts into commands for the drone.

Gad has been developing this project for a little more than a year, working on the flight portion for the last month. Sessions take 2 or 3 hours, and the drone’s battery lasts 15 minutes.

“Before I even started using it myself, I couldn’t believe I could control the drone with my head,” Gad said. “The first time that I moved the drone with my brain, it was remarkable, fascinating.”

Under the guidance of Zhuming Bi, professor of mechanical engineering and a Fulbright-Nokia Distinguished Chair in Information and Communications Technologies for the 2023–24 academic year, Gad and recent PFW graduate Hussein Bilal, M.S. ’23, are working to expand humanity’s possibilities. Similar to Gad operating a small drone, Bilal can manipulate a mechanical arm, twisting the device in directions using the power of his mind—and a brain-computer interface.

Teaching at PFW for 15 years, Bi continually pushes forward on human-machine interaction by acquiring and interpreting brain signals and directly communicating with a machine for enhanced controls.

“Traditionally, as far as working with robots, people just look at how robots can automate operations to release humans,” Bi said, “while recently, people find that on many occasions, it will be beneficial greatly by including themselves in the loop, and this requires to advance human-machine interaction.”

Bi’s goal is to develop a brain-computer interface to fuse human and machine intelligence to enhance the adaptability, reliability, and resilience of an intelligent system. He points out the great potential of brain-computer interfaces, or BCIs, in defense, space exploration, manufacturing, and healthcare. For example, a patient who loses a limb might control a robot via a BCI in performing daily activities without a caregiver’s assistance.

As colleague Don Mueller, associate professor of mechanical engineering said, Bi’s work is about how advanced technologies can be fully exploited to solve real-world problems and how those technologies can benefit society and improve the quality of life. Also, by having Gad and Bilal work on projects like these, Bi is developing the next generation of innovative researchers.

Learning to write code that allowed his computer program to access commands from his interface and connect to the robotic arm was a painstaking effort, Bilal said. Every piece of inputted data had to be perfect.

More difficult, Bilal had to learn about the robotic arm from scratch and how to control his thoughts to convey the commands he wanted to execute.

“It’s all about the interface,” Bilal said. “It’s not about controlling the robotic arm; it’s about controlling the smart machine. The system understands based on my brain signals, `Now we need to do this output.’ Your brain will produce specific signals, and based on that signal, you’ll have a corresponding application.”

The manufacturer does not supply the mechanics behind the device, Bilal said, so he had to learn those. Also, commercially available headsets allow four brain actions when he needed at least 12 to manipulate the arm. Some researchers have added facial expressions or muscle control to their efforts, but Bilal was determined to use only brain signals.

He took the four brain actions—A, B, C, and D—and added them together to create options such as AAAA or AAAB. Essentially, he reworked the system to create 256 combinations, and new possibilities.

“I was sitting, and once it started moving, I couldn’t actually believe it worked,” Bilal said. “I sent the video to my family because they saw how much I struggled and worked on it. I felt like I was actually going to graduate.”

Gad, who’s from Egypt, and Bilal, from Lebanon, were colleagues at Kuwait University before coming to PFW a semester apart.

Currently, Gad said, manual computer operators are quicker than brain-interface options, but he believes he can eventually equal the speed and become more efficient with fewer mistakes.

A video demonstration of Gad’s work will kick off PFW’s observance of Engineers Week in the ETCS Building lobby from noon to 1:30 p.m. on Monday. There will also be demonstrations from the Department of Computer Science on Tuesday, the School of Polytechnic on Thursday, and the Department of Electrical and Computer Engineering on Friday.