Six Teams Receive Funding to Tackle Innovation in Neuromotor Control and Ethics

What would you do with the ability to control a computer with just muscle signals — without having to learn to type on a keyboard, navigate with a mouse, or tap across a touchscreen? At Reality Labs, we’re investing in wristworn controllers like Meta Neural Band to create interfaces between humans and computers that are less robotic, more intuitive, and more inclusive.

The frontier of this research uses surface electromyography (sEMG) wristbands, which detect electrical signals sent by the muscles in the wrist and hand, and translate them into digital commands. We’re developing that wristband technology to control computer systems, starting with AI glasses like Meta Ray-Ban Display glasses and Orion, our true AR glasses prototype. It’s an area of research that could revolutionize the way we interface with all sorts of technologies, offering multiple degrees of freedom of control, and giving people access to more inclusive and intuitive control schemes.

Building on our ongoing commitment to advance responsible innovation sEMG strategies, and to deepen our collaboration with academia, our EMG Foundational Research team launched a request for research proposals (RFP) last summer. And after receiving more than 70 submissions from institutions across the globe, today we’re excited to announce the six award recipients: University of Central Florida; University of South Florida; University of California, Davis; Newcastle University; University of British Columbia; and Northwestern University. Each team was awarded $150,000 for their research, for a total of $900,000.

Goals of the Proposal Call

How do people learn new sEMG-based controls, and how can onboarding be streamlined? These questions were a main focus of the RFP. Teams were encouraged to identify approaches that could accelerate the learning curve for sEMG-based interactions and develop new methods for rich communication with computing systems.

We sought proposals encouraging innovation in sEMG controls that required collaboration between science and ethics-centric teams. Research projects explore this combination through diverse methods: user surveys on the learning process, embedding ethicists directly in the research, evaluating user sentiment and value, and identifying barriers to adoption.

We’re honored to present the six winning proposals below.

sEMG-Talk: Adaptive Interface for sEMG Control of Vocal Expression

These researchers will develop sEMG-Talk, an interface that combines sEMG and machine learning to generate speech, allowing its users to speak without the use of the mouth and throat. sEMG-Talk will make use of a device worn on the forearm that includes sensors to control a virtual vocal tract model, with potential applications in assistive technology, digital music creation, and gaming. The research will also include building out a new taxonomy of neuroethical considerations for the unique interaction model, which will inform experimental design throughout the project.

Learning sEMG-Based Control Across the Lifespan: Muscle Synergies and Individualized Feedback for Low-Movement Interactions

This work focuses on comparing different methods of teaching the transition between biomimetic and non-biomimetic controls for sEMG systems. The researchers will compare gamified, implicit feedback against step-by-step instructions in young and old cohorts who may have different preferences in learning styles. The team will also survey users to identify the ways different elements, such as sentiment and social support, affect the learning experience, so the team can work adaptively with users and find methods to promote user agency during learning.

From Silent Signals to Voluntary Control: A New Pedagogy of sEMG-Based Motor Learning

These researchers will evaluate how people learn to gain voluntary control over subtle muscle signals. The system reinforces activity in a target muscle to help individuals improve their ability to control it over time as they go about their day. The project will culminate in an evaluation with stroke survivors to explore how the system may help encourage activity in underused muscles. The ethics component of the project examines how this type of interface can maintain user agency, preserve control, respect individual preferences, and build trust between people and the systems they use.

Optimizing Multi-sEMG-Based Communication Bandwidth

The team will teach volunteers how to use sEMG during low-movement controls to increase communication capabilities with computing systems, while also coexisting with natural hand use. This paradigm could allow human-computer interactions that don’t interfere with daily tasks, and fit more naturally within everyday manual activity. Alongside this work, the team will survey participants on their perspectives on large-scale data collection with the aim of developing an ethics toolkit for future work. This strand will explore attitudes toward learning to use sEMG technology in general, and work to get clarity on any barriers to its adoption.

Long-Term sEMG-Based Interaction Facilitation Using Manifold Alignment and Kalman-Filter-Based Symbiosis

The UCF team will examine how people learn new motor skills in digital environments, research that may help improve technologies such as prosthetic limb control. Their approach uses co-adaptive, gamified training designed to allow both the algorithm and the user to improve coordination over time through continued interaction. An ethics lead will also be embedded directly within the engineering lab to help shape experimental design, examining potential privacy risks in sEMG data and assessing how the co-adaptive system may influence participants’ sense of agency and embodiment when using prosthetic technologies.

Optimizing Myoelectric Interface Learning in People with Intact and Injured Central Nervous Systems

This proposal focuses on evaluating how users go from low-bandwidth to high-bandwidth sEMG control, proposing to compare gradual shaping of control—adding muscle degrees of freedom sequentially—to AI-enhanced, co-adaptive shaping that trains multiple degrees of freedom at the same time. The team will work with people who have experienced a stroke or spinal cord injury as well as research participants who have not experienced an injury. The project will include embedded ethicists, an ethics advisory panel, and focus groups to answer questions related to the treatment of user data and the technology’s inclusivity.

We’d like to thank all the researchers who submitted thoughtful proposals and offer our congratulations to the winners! We’re excited to see the results of this research and the ways it can help make sEMG technology more expressive, inclusive, and responsible in the future.