T. Luczak, R. Burch, P. Nelsen, C. Freeman, H. Chander, J. Ball, J.A. Jones, J. Barlow, D. Saucier, M. Duclos, S. Grice, M. Taquino
Mississippi State University,
Keywords: machine learning, extended reality, wearable sensors
Summary:The Athlete engineering BaseLine Ecosystem (AeBLE) is a collection of integrated innovations impacting human performance, health, and wellness. AeBLE integrates artificial intelligence-driven mobile markerless motion capture, embedded wearable sensors, and augmented and virtual immersive technologies to train healthy movements and mitigate at-risk motion patterns which can lead to musculoskeletal injuries. Humans train and perform in many open and closed kinetic chain activities, often making unconscious modifications to their movement based on their center of mass, situational environment, object use, clothing and shoes, and ground interactions. Over time, altered movement patterns put undue stress at the weakest links—hands-wrists, feet-ankles, and lower back—resulting in injuries or diminished performance. AeBLE will provide users with information regarding: 1) body movements, 2) improvements from past training and performances, and 3) areas of potential risk for injury. At the foundation of the AeBLE technology is an accurate markerless motion capture smart device app. The AeBLE team is conducting machine learning (ML) in computer vision (CV) research using published and novel techniques and methods to create three-dimensional proprietary human pose datasets and human pose estimation models using two-dimensional video. Embedded with patent publication, “Wearable Flexible Sensor Motion Capture System” (#US20200008745A1), AeBLE can capture and infer kinematic movement and kinetic forces that occur during training and performance. Collecting, measuring, and assessing human movement will allow AeBLE to characterize, categorize, and individualize tendencies and forces over time that can lead to improved performance and reduction of musculoskeletal disorders. Use cases for AeBLE are widespread and can be defined by the individual, task, and activity. Using “athlete” personas to define a variety of users, including Sport, Industrial, Tactical, and At-Risk (injured) athletes, allows AeBLE to be modified to the situational environments where human athletes perform their tasks, train, compete, or encounter everyday activities. AeBLE can be used as a pre-habilitative platform to improve training and performance or to bridge the gap when an athlete finds themselves in the At-Risk athlete status by providing rehabilitation monitoring and validation. AeBLE is being developed by coordinated research and development (R&D) teams with expertise in ML, CV, data science, cognitive engineering, human and sports performance, augmented and virtual realities, physical therapy, technology innovations, wearable sensors, human factors, and other hardware and software development integrations. AeBLE is a work-in-progress project conducted by Mississippi State University (MSU) R&D teams comprised of MSU faculty, staff, students, athletic coaches, and practitioners. Several MSU R&D team members work in multiple academic departments which aids development of AeBLE across disciplines. Partners include members from Fortune 500 and regional industries and researchers from national and international research centers, as well as athletic trainers, strength and conditioning coaches, and sports scientists from teams in the National Collegiate Athletic Association (NCAA), the National Football League (NFL), the National Basketball Association (NBA), and Major League Baseball (MLB). Knowledge learned from our partners has guided our work to help solve their immediate problems by delivering laboratory-quality data, innovations, and guidance to all types of athlete personas.