Page 65 - OHS, October 2023

Gorodenkoff/Shutterstock.com all, and create an algorithm that can out- put personalized safety recommendations based on real-time data inputs. Over a period of two years, we built our own massive dataset, recording more than 10,000 unique movements made by hun- dreds of men and women of varying ages, body types, stress levels and injury histories. What we found is that every body truly is unique. There is no ideal formula for safety for male workers or for female workers, for old or for young, there are only safety insights for each individual and the way they are built and tend to move. We took this new dataset and were able to formulate it into a universal measure- ment we call Movement Intensity, based on high-frequency Inertial Measurement Unit (IMU) data. Put most simply, Movement Intensity measures how difficult a movement is for you to make. It accounts for factors like the jerkiness of your movement, the size of the load you’re carrying or the angle at which you’re doing so, and even physical condi- tions that can impact how ready you are to move that way – did you sleep well the night before, are you feeling extra fatigued or stressed, or is an illness or disease impacting your ability to move. Two people could each lift a box of the same size and weight and come out with significantly different Move- ment Intensities, because they each have unique factors to their body, impacting how difficult the movement is for them to make. Using this Movement Intensity data, we trained a neural network to process over 100 different elements of individual movement patterns. We’ve then applied the neural network to process the data our wearables collect in real time. As a result, workers using our wearables at partner companies like IKEA and DHL are get- ting real-time custom alerts, letting them know when they’re making a movement that’s strenuous for their unique body (not for 75 percent of males), and are putting themselves at greater risk for injury so they can stop it before it happens. The ROI for Inclusive Safety This personalized approach isn’t just bet- ter for workers. Of course, reducing time off due to injuries and workers’ compen- sation claims impacts the bottom line, but the data custom safety tools like wearables provide today, also inform other produc- tivity factors for managers. Understanding who in the facility has trouble bending safely, and who is able to output at high levels without much risk of injury can directly inform the job functions they should hold to maximize overall shift performance. Data finding that everyone is moving safely consistently while reaching high outputs, or alternatively, trends in un- safe movements, could directly correlate to the quota or bonus threshold you’re using which could be adjusted to optimize high performance against reduced injuries. For too long there’s been a perception that safety and performance are inher- ently at odds. In reality, leveraging data to inform safety actually allows us to better understand where that boundary lies be- tween maximized performance and mini- mized injuries. Matthew Hart is the Founder and CEO of Soter Analytics. Hart, along with co- founder Alexey Pavlenko, originated the idea for Soter Analytics after working in the mining industry and developing an interest in sensor technology. REFERENCE 1. www.commerce.gov/news/blog/2022/10/ manufacturing-opens-more-doors-women www.ohsonline.com OCTOBER 2023 | Occupational Health & Safety 65 See us at NSC, Booth #2208