Data mining is now increasingly being applied in occupational health and safety management. Organizations are looking to their leading indicators to, essentially, predict where the next incident will occur and hopefully prevent it. Safety solutions providers are now also venturing into data analytics in their march towards integrated OHS products and services offerings.
Organizations with a fairly decent OHS management system use large amounts of data to measure performance and identify opportunities for safety improvement. An increased number of musculoskeletal injuries among employees, for example, may indicate a need to focus on ergonomics.
In recent years, leading indicators have become popular among safety managers as a valuable metric for prevention efforts. These are things like reports on near misses, job hazard analysis, observations, safety inspections, training completion and effectiveness, and even leadership engagement activities. These metrics, according to data experts, feed well into the analytics model for predicting and preventing incidents or injuries.
If you’re not quite sold on the idea of using data to prevent injuries, consider this recent experiment: The Language Technologies Institute at Carnegie Mellon University (CMU) in Pittsburgh has collaborated with Predictive Solutions, a provider of occupational safety software based in Oakdale, Pa., to develop several predictive models that can predict safety incidents. The study used leading and lagging indicators from actual workplace data across 250 work sites. The result was an accuracy rate of between 80 and 97 per cent in predicting incidents.
The Language Technologies Institute is the same CMU department that helped IBM develop the Watson supercomputer that is now helping doctors diagnose rare diseases.
It’s always a good thing when a piece of new technology expands its application beyond the geeky, into the real world and for the greater good.
Safety professionals are trained to look at data in the course of their duties — for legal compliance, workers’ compensation reporting and performance indicators. With their data analytical skills and the advancements in analytics technology, I predict a safer future for all workers.
Organizations with a fairly decent OHS management system use large amounts of data to measure performance and identify opportunities for safety improvement. An increased number of musculoskeletal injuries among employees, for example, may indicate a need to focus on ergonomics.
In recent years, leading indicators have become popular among safety managers as a valuable metric for prevention efforts. These are things like reports on near misses, job hazard analysis, observations, safety inspections, training completion and effectiveness, and even leadership engagement activities. These metrics, according to data experts, feed well into the analytics model for predicting and preventing incidents or injuries.
If you’re not quite sold on the idea of using data to prevent injuries, consider this recent experiment: The Language Technologies Institute at Carnegie Mellon University (CMU) in Pittsburgh has collaborated with Predictive Solutions, a provider of occupational safety software based in Oakdale, Pa., to develop several predictive models that can predict safety incidents. The study used leading and lagging indicators from actual workplace data across 250 work sites. The result was an accuracy rate of between 80 and 97 per cent in predicting incidents.
The Language Technologies Institute is the same CMU department that helped IBM develop the Watson supercomputer that is now helping doctors diagnose rare diseases.
It’s always a good thing when a piece of new technology expands its application beyond the geeky, into the real world and for the greater good.
Safety professionals are trained to look at data in the course of their duties — for legal compliance, workers’ compensation reporting and performance indicators. With their data analytical skills and the advancements in analytics technology, I predict a safer future for all workers.
