At its core, safety is an information-based discipline. This information can take many forms; it could be leading and lagging safety performance indicators, emissions numbers, risks noted over multiple observations at the shop floor, audit logs and so on. Today’s unfortunate reality is that environment, health and safety (EHS) leaders often spend more time worrying about the cleanliness, completeness and comprehensiveness of that information than they do thinking about how to strategically leverage the insights embedded within it.
Fortunately, artificial intelligence (AI) is changing the narrative. Stressed that your employees aren’t adequately documenting near misses or observations? AI can help. Struggling to stay up to date on the latest regulatory developments and ensure 100% compliance? AI can help. Straining to identify relationships within EHS datasets that will unlock performance gains? You guessed it—AI is here to help.
Broadly speaking, AI is allowing EHS professionals to spend more time doing what they got into this business to do: tactically implement programs, processes and procedures that help every worker get home safe every day. Let’s explore some of these use cases in detail.
See Everything That’s Happening
EHS professionals can't be everywhere at once. Walking the floor is often in the daily routine, but even then, there will be safety issues that will only be documented if a frontline employee takes it upon themselves to report it. With AI, closed-circuit television (CCTV) cameras can be trained to automatically watch for, detect, and report near misses and policy nonconformances. This can include speeding vehicles, employees incorrectly wearing personal protective equipment (PPE), incursions into exclusion zones and much more.
The beauty of AI for this use case is that EHS leaders can train it to see whatever they want to look for. Safety leaders can be virtually omnipresent in a way that wasn’t previously possible.
Finding Value in Large Datasets
Organizations with tens or hundreds of thousands of employees generate a lot of safety data. Observations, incident reports, audit findings, inspection logs, and the list goes on and on. Filtering through the noise is difficult, but AI can browse these datasets in a fraction of the time it would take a human.
One thousand incident reports in different languages, formats or styles can be sorted into 10 high-level categories with a click of a button. Countless unstructured observation descriptions can be boiled down to a few key themes in seconds. AI presents major opportunities for time savings and efficiency gains.
Predictive and Prescriptive Analytics
Since AI is well-versed in looking at data, it’s great at finding meaningful relationships or correlations. That means if the circumstances that previously led to a negative outcome (compliance violations, serious injuries and fatalities, etc.) reoccur, AI can automatically flag it. AI can then let the EHS leader act themselves or recommend an action that might help reduce current risk levels.
For example, if there was a serious hand injury, AI could automatically suggest refresher courses on machine guarding and PPE to every user affiliated with a task where a job safety analysis includes risk of laceration. This idea isn’t novel in EHS, nor are the solutions for it; however, the innovation is in how AI makes EHS management more accessible to many more organizations.
Stay on top of Regulations
Ensuring compliance with evolving regulations is a complex task for organizations. Machine learning algorithms continuously analyze regulatory updates from every conceivable source to ensure that EHS management systems are always up to date. AI can take it a step further and automate the generation of compliance reports, further reducing the administrative burden on organizations and minimizing oversights.
Improve Training and Drive Engagement
Mature EHS programs are built on countless documents, policies and procedures. You want employees to know them like the back of their hand, but they have jobs to do and might interact with them sparingly. What if AI could take key documents and automatically create micro lessons to help employees?
Well, it can! Instead of an EHS leader spending weeks or months producing training programs, AI can do it in minutes. Instead, the now unburdened EHS manager can use that time to talk with frontline employees and build relationships.
Challenges with AI
Despite the evident advantages of AI, many organizations remain skeptical about its capacity to comprehend, interpret and organize critical information that drives performance and ensures compliance. A significant portion of EHS information is highly sensitive for a variety of reasons, so trepidation about data privacy, security and general effectiveness is commonplace.
Equally critical is addressing concerns about bias and equity. AI models are built by people, and people have their own thought processes and ways of doing things. If not accounted for, AI models become a reflection of the person or people that built them—biases, blind spots and all. Proactive efforts are essential to identify and mitigate biases in AI models, ensuring outcomes that achieve the goal of improving performance to the maximum extent possible.
Navigating regulatory compliance poses another challenge in the application of AI in EHS. Are you ready to put your compliance program in the hands of a faceless, nameless, liability-free stack of code? What if it misses something and your organization must pay thousands or millions to correct the error?
There are plenty of issues to watch out for, but a proactive strategy for utilizing AI in EHS can help your organization harness its power with limited downsides. When implementing AI tools, be sure to do the following:
- Regularly monitor outputs and results. AI models need to be “taught,” so check their work just like you would that of a student.
- Clarify who has final say. If AI suggests one thing and an organizational leader suggests another, which suggestion wins? Avoid conflict down the line by clearly articulating roles and responsibilities up front.
- Assess liability. Especially important in regulatory use cases, be sure to review limitations and legal responsibility before leaning on a virtual tool for real-world compliance.
- Determine which AI use cases will give you the best ROI. AI can do a lot, but that doesn’t mean your organization needs all that functionality. What problems do you need help with? What AI application will make an immediate impact? Address those first.
- Understand the tools you chose to use. Not all AI tools are created equal. They’ll have different underlying methodologies, different security profiles and different performance parameters, to name a few. When implementing an AI tool, especially for a new use case, do your homework so you know what you’re getting and how it works.
- Talk to your IT team. This one goes without saying, but bring in the most technologically-literate people to conversations about emerging technology.
The integration of AI in EHS applications is reshaping the landscape of workplace safety. AI-driven EHS solutions will empower organizations to be more proactive by enabling better risk mitigation, real-time monitoring and streamlined compliance processes. AI will be less of a tool and more of a partner in the never-ending quest for zero harm.
The possibilities of AI are only now coming into full view. As we harness AI’s potential, the future of EHS looks promising—one informed decision at a time.
Trevor Bronson is director of portfolio strategy with Intelex, a provider of SaaS-based environmental, health, safety and quality (EHSQ) management software.
