June 23, 2026 — Southwest Research Institute (SwRI) and St. Mary’s University are collaborating to expand and validate a system to estimate metabolic cost prediction for evaluating movement efficiency to improve performance or rehabilitation. The team will use ENABLE™, an SwRI-developed markerless motion capture technology, musculoskeletal modeling, and machine learning to improve metabolic cost prediction for healthcare. The project is supported by a \$127,750 grant from the St. Mary’s-SwRI Technology & Applied Research (S2TAR) program, which fosters collaborations between researchers from both organizations.
“Metabolic cost is the amount of energy a person’s body uses to do an activity, like walking or running,” said Dr. Nicholas Vandenberg, a research engineer in SwRI’s Mechanical Engineering Division and co-principal investigator of the project. “A common secondary objective of rehabilitation can be targeting the efficiency of retrained movement patterns. Being able to reliably estimate metabolic energy expenditure could provide an important tool for rehabilitative specialists.”
Vandenberg and Dr. Ricardo Ramirez, an assistant professor of electrical engineering at St. Mary’s, will build on Ramirez’s work using 2D video footage and machine learning algorithms to predict metabolic costs, and extend their approach to 3D video using the SwRI-developed Engine for Automatic Biomechanical Evaluation (ENABLE) markerless motion capture tool.
Markerless motion capture leverages computer vision algorithms to collect 3D motion data for biomechanical analysis in research, clinical, and sports science applications without the need to attach physical body markers to human subjects. ENABLE combines SwRI’s biomechanical modeling expertise with computer vision and deep learning to develop algorithms that enable accurate, reliable, markerless motion capture.
SwRI and St. Mary’s will use ENABLE to capture movement data of people with below-the-knee amputations and people without mobility-related limb loss. The researchers will use a metabolic cart to directly measure how much energy their bodies are using. The team will then run the 3D motion data through musculoskeletal models with metabolic probes to validate the machine learning predictions. By comparing model outputs and experimental metabolic measurements to machine learning predictions, the researchers will be able to validate and refine the algorithms.
“We hope this work will improve the quality of life for people with mobility challenges by helping clinicians and researchers better understand how individuals move and where their bodies are working harder than necessary,” Ramirez said. “By uncovering subtle gait mechanics and inefficiencies, especially in people who use prosthetic devices, this tool could support more personalized rehabilitation, better prosthetic design and fitting, and ultimately reduce fatigue and discomfort during everyday activities.”
The researchers will use OpenSim models, which include individual muscle fibers and allow them to break down metabolic cost by various muscle groups. This could be useful as additional machine learning training data.
“Our eventual goal is to be able to predict the metabolic costs for a full range of activities and how they are distributed across muscle groups. These data can in turn validate and improve the machine learning algorithm,” Vandenberg said. “This could have a remarkable impact on athletic training and rehabilitation.”
This project was funded (fully or in-part) by St. Mary’s University and Southwest Research Institute.
For more information, visit ENABLE or contact Joanna Quintanilla, +1 210 522 2073, Communications Department, Southwest Research Institute, 6220 Culebra Road, San Antonio, TX 78238-5166.