Workplace safety is a tremendous responsibility. According to 2017 numbers from Liberty Mutual, U.S. companies spend the equivalent of $1 billion every week, combined, on non-fatal injuries in the workplace.
The good news is, keeping personnel and assets safe is becoming an increasingly manageable challenge, thanks to modern technologies — including big data and predictive analytical tools.
What Kinds of Data Does This Require?
If the occupational health and safety industry is to modernize itself, it needs to take advantage of as many kinds of data as possible. We’re fond of saying businesses must run like well-oiled machines because it’s true. But even the most innovative companies rely on fairly repeatable and routine tasks when you distill things to the basics.
In short, companies have a variety of data types to draw from that can be useful in anticipating and mitigating risks and otherwise informing safety policy:
- Rates and details of reported safety incidents
- Whether or not the company is within safety training compliance
- Reported maintenance items and remediation actions taken
- Costs for previous injuries and accidents
- Closer to real-time information on machine condition and asset locations
- Employee experience levels and self-reported health conditions
- Data from condition-monitoring wearables
- Environmental data and facility telemetry
So how do we put all this into practice? Physical and stress-related health and safety issues are common in offices and warehouses alike. What are some of the ways technology can help keep us safer and better attuned to workplace risks?
How Companies Put Safety Data to Work
In the most practical terms, what do we hope to accomplish with this data? Ultimately, the goal is to use analytics to build predictive models of things like:
- Based on their job functions, which workers are likeliest to sustain certain types of injuries?
- During which times of day are certain types of incidents more frequent?
- Which areas, tools, and equipment types are the biggest drivers of safety incidents?
- How much do injuries cost your company every year, and are there investments worth making to bring that number down?
In a way, that last point is the key to all this. As employers in every industry begin taking a more scientific, data-driven look at their incident, accident and compliance metrics, they’re likely to find areas where the cost of the risk in continuing operating in a certain way, or with a specific type of equipment, doesn’t make financial sense anymore.
In other words, companies can use predictive modeling to learn where injuries tend to occur and are likely to happen in the future and bring automation into areas where the costs to deal with employee injuries over time are higher than the price of an equipment upgrade. Companies can use predictive models to weigh expenses against existing levels of risk. If a company wants to eliminate injuries in cross-traffic areas where workers transport bulk merchandise across the warehouse, it’s probably time to explore the ROI of automated pallet trucks for warehouses, or something similar.
There is also the legal side of things to consider. We mentioned wearable health and safety devices above — and it’s because lots of employers are taking a serious look at this type of technology to keep a closer to real-time eye on employees as they navigate or work in dangerous conditions. Several professions have a reputation for placing employees at an unusually high risk of knee injuries, for instance. Having access to information about employee behaviors or work techniques that might heighten that risk could be a useful prevention tool.
On the one hand, we’re talking about wearable devices that collect data such as location, posture, heart rate, stress level, temperature and more — and that’s a big can of legal worms. On the other hand, all this incoming data makes it easier for managers, safety officers, and building inspectors to predict incidents and to potentially trace the cause if an accident does occur. And if a workplace injury case must escalate to a court-type situation, having access to operational data and being able to prove you took reasonable precautions based on predictive risk models goes a long way toward keeping your name clear.
Predictive Analytics: Better With Friends
We’ve been looking at the ways in-house data can help companies weigh the cost of equipment upgrades against the likelihood of future incidents. Third-party data also has a role to play here. Predictive analytics for safety can also fold in data from sources like:
- Workforce management data and scheduling tools
- Data from peers, vendors, and partners
- Weather information
- Market trends
- Data from badges and authentication systems
In other words, planning an enterprise around labor shortages, weather events, problems in the market, material shortages and data from security systems have plenty of implications for workplace safety. Predicting when a bad storm can threaten to take your facility offline gives you time to plan for contingencies, up to and including installing backup equipment and drawing up procedures for getting your people home safely. As far as market disruptions go, enterprise planning tools with predictive analytics can help companies plan for incoming demand, staff up as necessary and ensure employees aren’t overworked and exhausted later on as they scramble to catch up or get ahead.
The most significant benefit of predictive modeling for businesses is that these tools only become better and better equipped over time to render meaningful recommendations. The more data sources you fold in and the more comprehensive your data program, the higher the number of safety insights you’ll be able to gather. Maybe most importantly, taking this proactive, predictive approach means your company or facility isn’t only “getting by” in between incidents and inspections — you’re actively weighing new data from business systems, industrial control devices, HR and more, and taking a fresh look at safety and compliance even as conditions change on the ground.
