PIs: Xiang, Yujang/Yang, James Proposals: 1700865/1703093
Currently available safety evaluation tools for manual material handling are based on static lifting conditions, such as the NIOSH lifting equation, which gives a subject's maximum lifting weight, distance and height for static lifting. Another popular method for evaluating injury risk for manual material handling is to measure the lumbar spine stress for back pain. However, these methods are all static and cannot accurately reflect dynamic motion generation and injury prevention. Using biomechanical models provides relatively new, alternative techniques that allow direct testing and individualized results. In this collaborative project, the PIs will develop a predictive lumbar spine musculoskeletal model for a dynamic manual material handling. The project aims to develop an efficient tool for injury risk assessment and prediction in dynamic lifting process. Project results will have potentially wide application in manual material handling ergonomics and will further advance lifting biomechanics. The project includes education opportunities for both undergraduate and graduate students. A two-week summer workshop will be offered to underserved minority middle and high school students at University of Alaska Fairbanks in collaboration with the Alaska Summer Research Academy. Another week-long summer workshop will be offered to local workers with lifting jobs and Hispanic middle school students at Texas Tech University.
The project focuses on developing an inverse dynamics optimization-based (predictive dynamics) method for dynamic manual material handling using a musculoskeletal model with dynamic strength data. Dynamic strength data will be developed for key joints, such as the knee, hip, ankle, elbow and lower spine. The tool developed for predicting injury events in dynamic lifting systems will advance the current offline procedure to an online, near real-time optimal motion control and injury prevention system. Injury is defined in two ways: by measuring lumbar spine compression and shear stresses using a musculoskeletal model, and, in joint space, by measuring the percentage of joint torque for key joints to predict injury. Specific objectives are to: 1) derive a general dynamic strength model and validate the model parameters from experiments; 2) introduce and experimentally validate a lumbar spine muscle model; and 3) implement these models with a nonlinear programming algorithm to optimize the dynamic lifting motion during manual material handling for minimum injury and experimentally demonstrate proof-of-concept. Muscle intra/inter-joint coupling will be modeled and the lumbar spine area will be added, thereby generating a musculoskeletal model to measure lumbar stresses for back pain in the dynamic lifting process. The dynamic strength parameters in joint space estimated from active human subjects will provide the model with improved accuracy and efficient calculations that can be used to evaluate injury for complex non-periodic functional tasks that may not be experimentally verifiable by traditional means. The model, which enables near real-time calculations of dynamic manual material handling as a function of time, can be used to establish an individual-specific dynamic limiting lifting process, as well as optimal strategy for lifting. Project results will serve as new guidelines for manual material handling ergonomic design, considering dynamic effects. In addition, better understanding of various constraints in optimization formulation inherent in the human physiological system, as well as motion adaptation to these constraints, will contribute to the field of human locomotion study and complement existing design principles for manual material handling.
