Stepping Towards Success with AI-Sole

For Damiano Zanotto, an Associate Professor of Mechanical Engineering at Stevens Institute of Technology, the inspiration for the research behind his entrepreneurial venture, AI-Sole, came from standing on his own two feet. Literally.

As a boy in Northeastern Italy, Zanotto’s feet would often cause him pain at the end of an active day.

“My fascination with biomechanics of walking and assistive technologies stems from a personal struggle—living with flat feet,” Zanotto said. “As a teenager, I worked summers in local fashion and textile factories as a warehouse operator. Carrying heavy loads was far more painful for me than for my coworkers because of overpronation and excessive stress at my ankle. This experience made me acutely aware of the impact of musculoskeletal conditions and gait impairments on daily life. It ultimately shaped my research focus on rehabilitation robotics during my doctoral studies.”

As he grew up and developed a strong interest in engineering and robotics, Zanotto hoped to find a solution that would help those with mobility challenges. “I enjoyed taking things apart to see how they worked,” he added. “I've always loved math and physics—especially classical mechanics—and I've been passionate about coding since middle school, so pursuing a degree in mechanical engineering with a concentration in robotics seemed like the perfect way to combine all these interests to tackle real-world challenges.”

Zanotto received his undergraduate degree in mechanical engineering from the University of Padua, the second oldest university in Italy, and began his PhD program there as well, specializing in mechatronics, an interdisciplinary field that integrates mechanical engineering, electrical engineering, computer science, and control systems.

“I was drawn to robotics,” Zanotto said. “In robotics there are three main areas of research—perception, reasoning, and acting. Essentially, how do robots perceive the environment, how do they understand it, and how do they leverage that understanding to act on it? My PhD research focused on action and perception—that’s what got me started on researching wearable sensors and devices.”

A collaborative research program with the University of Delaware (UD) offered the opportunity for Zanotto’s first trip to the United States, in 2010. “I got a scholarship to conduct research there for a year,” he said. “During the last year of my PhD program, my UD advisor invited me to stay longer for a postdoc.”

Zanotto’s research at UD involved developing robotics to aid physical therapists doing rehabilitation work with patients whose movement was impaired. In 2013, Zanotto began a new position at Columbia University as a research scientist. While there, he went thought a bootcamp run by Columbia Biomedical Engineering. “It’s somewhat like I-Corps, but it’s a semester-long intensive entrepreneurship training program, focused on commercialization of healthcare-related technology,” he said.

He was also accepted into the Columbia Engineering Translational Fellowship Program (TFP), which offers a year of salary support to help maximize opportunities to commercialize research. Zanotto used his TFP funding to further pursue commercialization of a smart footwear device.

“At UD and then at Columbia Medical Center, talking to physical therapists and neurologists, I was surprised to learn that they tested mobility just using a stopwatch and a score sheet,” Zanotto said. “The most common protocols they use are older than I am and had not been improved upon or changed for a long time.”

“How can we use technology to assess mobility in a better way?” Zanotto asked. “That’s the question that pushed me to explore wearable technology. As engineers, I believe we have an obligation to apply the tech we develop in the real world; it’s pointless to develop it just for the sake of studying a new sensor or algorithm.”

Walking may appear to be a simple motor task, but maintaining stability during walking requires the coordination of many lower-body muscles. For this reason, conditions that affect the nervous system's ability to sense, plan, or execute motions often impact the way an individual walks. “If you can monitor the way someone walks, you can potentially diagnose and manage conditions more accurately,” said Zanotto. “We’ve focused on addressing neurological conditions that progress over time.” These include Duchenne Muscular Dystrophy, Spinal Muscular Atrophy, Parkinson’s disease, and others.

“Our intent is to capture a disease’s progress and severity with a higher level of sensitivity,” Zanotto said. “Measuring with our innovation allows for a more accurate tracing of disease progression and provides a more sensitive assessment of the effectiveness of new treatments.”  

In 2016, Zanotto started his own lab at the Stevens Institute of Technology. Along with teaching, he continued to pursue the idea of smart footwear as a line of research. Zanotto, along with a team of researchers from his lab, first participated in a NY I-Corps Hub Regional I-Corps in Winter 2023. In the next season, the team participated in the NY I-Corps Hub Spring 2023 National I-Corps.  

“Going through regional and national I-Corps was important; it allowed us to fine-tune the beachhead market,” Zanotto said. “We spoke to 128 stakeholders during customer discovery and determined that there was a great opportunity with pharmaceutical companies conducting drug discovery trials. They really value precision and accuracy, and our technology could help measure performance differences with greater granularity and sensitivity. Doing so enables these companies to capture treatment effectiveness with smaller sample sizes and shorter trials, reducing overall trial costs.”

The established protocol for measuring motor function is at the doctor’s office. However, those assessments can only take place during such a visit, which might last 20 minutes. That snapshot doesn’t capture the whole picture or necessarily reveal real-world performance.

“Typical wearables in clinical research do not provide granular detail,” Zanotto continued, “Because ours is embedded in the footwear, we can capture broader parameters and target different ‘gait domains’—different aspects of how a person moves.”

Smart footwear generally includes devices that can measure temporal, spatial, and kinetic parameters, for example, how many steps are taken over time, the ratio between phases of the gait cycle, how weight transfers from one leg to the other, and foot-ground clearance, among others.

“We are the first technology capable of measuring all of these parameters in everyday settings,” said Zanotto. “There are many potential applications, but we are focused first on the clinical application, looking at walking in individuals with movement disorders, primarily targeting neurological conditions

Since 2019, Zanotto has secured more than $2.97 million in grants—primarily from the National Science Foundation (NSF) and National Institutes of Health (NIH), and from the NJ Department of Health with some nonprofit foundations. Zanotto has also utilized numerous New York City Innovation Hot Spot resources, including the workshop series about intellectual property and how to secure non-dilutive funding from federal agencies, including the SBIR and STTR grants from the NSF and Department of Defense (DoD).

“NSF and NIH grants are great for helping researchers in academia carry out fundamental and translational research, but those grants don’t typically provide resources for prototype development,” added Zanotto. “That’s where we really took advantage of the New York City Innovation Hot Spot prototype program to advance our work.”

“With the support from that program, we were able to achieve a forty percent reduction in weight and double the battery life,” he said. “We transitioned from a bulky and heavy wearable to a 3-D printed insole that is very thin and lightweight,” he said. “It can measure data for 12 hours per day.”

The AI-Sole team comprises a small group of researchers from Zanotto’s lab. The team plans to apply for an SBIR by this summer and hopes to incorporate then.

“Our biggest challenge so far has been trying to cross the classic startup ‘Valley of Death,’” Zanotto laughed. “We recently secured a $2 million NIH grant in collaboration with Columbia University Irving Medical Center for a clinical study on neuromuscular disease. The funding supports the identification of new digital biomarkers. The real challenge is attracting serious investment to help us transition our 3D-printed insole into a production-ready prototype.”

Along with running his lab, Zanotto teaches several courses related to robotics: an undergraduate controls course for mechanical engineering majors, and two graduate courses on engineering dynamics and wearable technologies. “My favorite course is the one I developed from scratch on what I enjoy most—developing wearable robotics and sensors,” he said. “We cover design and testing of wearable robots for assistance and rehabilitation as well as the sensors and algorithms used in wearable-based motion analysis.”

 “That course has become a great channel for attracting talented and passionate students to my lab!,” Zanotto added. They take the course and then they enroll on a thesis or research project with my lab. Today, nearly all research projects in my laboratory at Stevens Institute are dedicated to developing technologies for individuals with impaired walking function. Our mission is to advance next-generation wearable technologies that enhance physical rehabilitation, improve human motion analysis, and ultimately, elevate quality of life.”

Zanotto also recently became a father for the first time. “That takes all of my spare time,” he said with a smile.