SBIR Phase II: Predicting Musculoskeletal Injury Risk of Material Handling Workers with Novel Wearable Devices

这项小型企业创新研究（SBIR）第二阶段项目的目标是证明，工业工人使用的独立的、腰带安装的互联网连接可穿戴设备可以检测高风险的起重活动，促进安全的起重实践和行为改变，并预测由于不安全的起重而导致的肌肉骨骼损伤的风险。每年有超过60万名工人因起重相关活动而遭受肌肉骨骼损伤，美国公司每年花费超过150亿美元。工伤影响员工士气、缺勤、生产力损失和员工流动，所有这些都是对公司有效运营的挑战，也是造成人类痛苦的不必要原因。通过开发一种可穿戴设备，可以检测高风险的起重活动并向工人提供即时反馈，可以促进更安全的起重实践，减少不安全的起重数量，从而减少伤害。该项目包括三个主要技术目标：i）开发机器学习算法，以从传感器数据中检测提升事件，并测量与那些提升事件相关联的风险相关度量。ii）在多个客户现场的工业环境中部署该设备12个月，工人执行的高风险提升次数随着时间的推移而量化，以衡量系统推动劳动力行为变化的能力;以及ii）基于所测量的风险度量来开发能够预测肌肉骨骼损伤的可能性的模型。预计该项目的成果将大大降低肌肉骨骼损伤的风险，为工业公司及其潜在客户保险公司的明确投资回报价值主张铺平道路。

This Small Business Innovation Research (SBIR) Phase II project has the objective of demonstrating that discrete, belt mounted internet-connected wearable devices used by industrial workers can detect high risk lifting activities, promote safe lifting practices and behavior change, and predict the risk of musculoskeletal injuries due to unsafe lifting. Each year over 600,000 workers suffer a musculoskeletal injury due to lifting related activities, which cost US companies over $15bn annually. Worker injuries affect employee morale, absenteeism, productivity loss and employee turnover, all of which are challenges to the efficient running of a company and are a unnecessary cause of human suffering. By developing a wearable device that can detect high risk lifting activity and provide immediate feedback to workers, safer lifting practices can be promoted and a reduction in the number of unsafe lifts registered, leading to a reduction in injuries.The project includes three main technical objectives: i) the development of machine learning algorithms to detect lifting events from sensor data, and to measure risk related metrics associated to those lifting events. When a lift is considered high risk, real-time feedback will be provided to the worker; ii) the deployment of the device in an industrial setting at several customer sites for 12 months, with the number of high risk lifts performed by workers quantified over time to measure the ability of the system to drive behavior change in the workforce; and ii) the development of a model that can predict the likelihood of musculoskeletal injures based on the risk metrics measured. It is expected that the outcomes of the project demonstrate a significant reduction in the risk of suffering musculoskeletal injuries, paving the way for a clear return on investment value proposition for the industrial companies and their insurance carriers who are potential customers.