7 min read
Opinions expressed by Entrepreneur contributors are their own.
Artificial intelligence (AI) is everywhere. Embedded into our everyday lives, from our ridesharing apps to the algorithms on our social media channels, AI has the potential to revolutionize every industry. However, there are a number of challenges that many technologies still need to overcome before they can actually implement AI — and that’s especially the case in the risk management industry.
Today, companies across every industry rely on environment, health and safety (EHS) procedures to promote a safer and more compliant workplace. Essential to companies’ risk management strategies, EHS programs are commonly used to help companies avoid unwanted events. Many hope that AI can help them understand how EHS events are impacting their bottom line. However, to date, predicting the complex root cause of an unwanted event has been difficult, if not impossible.
For businesses to truly leverage the power of AI in EHS platforms, they first need to ensure they have the infrastructure in place to implement integrated risk management data strategies. These must extend across each business unit to clearly understand the risk factors that could cause an incident. Through a comprehensive view of their risk environment, businesses can better leverage EHS platforms and move closer to utilizing AI to predict unwanted events.
Going back to the basics of EHS management
Developing a strong EHS strategy is essential. As the foundation of companies’ risk management strategy, EHS programs help companies prevent workplace incidents that could negatively impact workers, business operations, the surrounding environment, communities, and physical assets.
To truly understand the importance of EHS management, let’s take a look at the British Petroleum (BP) Deepwater Horizon oil spill. In April of 2010, an oil drilling rig exploded, causing the largest oil spill in the history of marine oil drilling operations. Releasing 134 million gallons of oil into the Gulf of Mexico for an 87 day period, the oil spill killed thousands of marine animals, contaminated habitats, and spoiled miles of coastline for animals and humans alike.
During the investigation, it was discovered that BP had made a series of operational decisions that dramatically increased risk in a number of critical areas. For example, BP employees testified that instead of considering their occupational safety in their performance evaluations, they were judged on how quickly they performed a task because tasks completed more quickly reduced the overall cost of a project.
This incident, and many others like it, showcases the importance of integrated operations that allow environmental, health and safety risk management to influence enterprise risk management. When organizations prioritize financial risk management over non-financial risk management, there’s a higher chance of significant, detrimental incidents occurring.
However, significant events like Deepwater Horizon also illustrate the collision and sometimes conflict between risk management practice and risk management information. As we know, there are massive amounts of data in business today. While that data exists, the challenge and opportunity is to leverage that data to make informed decisions and to combine the merits of risk management data with an organization’s risk culture and decision-making.
According to the Disaster Recovery Journal, 64% of organizations in 2020 were still using ad hoc tools and methods to manage business continuity and risk management. This traditional approach doesn’t provide the speed, scale and agility needed to keep up with today’s business environment. Businesses
need quick and efficient access to integrated operational and enterprise risk data across areas like health, safety, environment, and social impact.
For businesses to harness AI, they first need a holistic understanding of risks across the entire business, not just within each business unit. It is the interconnectivity between the people, processes, events, departments, and controls within an organization that provides the deeper insight necessary to identify risks, and thus to prevent unwanted outcomes proactively. Once this infrastructure is in place, companies can begin to think about how AI can play a role in their risk management strategy.
Breaking down the siloed nature of EHS management
Despite the fact that many risks cut across disciplines — such as environment, health, safety, social, and quality control — risk management often operates in silos, making a holistic understanding of risk indicators and consequences difficult. Traditionally, companies conduct risk assessments every six to 12 months and then reassess those risks periodically to create the next year’s risk register. However, due to how quickly new risks evolve, businesses need to break down their risk management silos to create a more agile approach.
In the past, the problem is that a lack of data has acted as a barrier to predicting unwanted events in the environmental, health and safety realm. In many cases, businesses overlook the historical data that could identify red flags leading to larger incidents. Operating in silos means that different parts of the business may not be communicating with one another to collect and analyze this historical risk data. Simply adding AI is not going to solve the problem. Without integrated, standardized and streamlined data across each business unit, AI can’t predict when something is about to go wrong.
For EHS technology to leverage AI, businesses need to focus on monitoring and management of controls. The core of risk management is standardizing the process of identifying, implementing, monitoring and managing controls. By streamlining the management of critical controls, businesses can ensure that every risk is identified, accessed, and controls are implemented, understood and effective.
By unifying the data, businesses can connect between enterprise and operational risk and between financial and non-financial risk. This central oversight allows organizations to break down the departmental and business silos that cause them to miss the control failures, resulting in unwanted events.
EHS software is often a catalyst towards standardizing control management, but an organization’s people play a big role. However, many organizations have found that a lack of employee buy-in on the technology causes the standardization process to crumble. The problem is that there’s often a skills gap or reluctance to change within the workforce that hinders the implementation of new technology. Businesses need to understand that people are more likely to use a new platform if they can clearly see its value and understand how to use it.
Educating employees on the goals of EHS software and how it can facilitate their day-to-day tasks and reporting will promote employee buy-in. Once they see and understand first-hand how real-time, accurate data capture and risk monitoring plays into their company’s larger risk management strategy, they will feel more empowered to embrace the system. It isn’t until employees start using EHS software properly that organizations can break down the silos of EHS management and connect the dots between leading indicators and incidents.
AI relies on consistent, standardized data to be able to recognize patterns and learn. Thus, until organizations have mature enough EHS programs to adopt technologies that collect, produce and report that data, the market is not ready to reap the benefits of AI. Right now, many EHS programs are still not aggregating data with the volume and consistency necessary to leverage AI.
So, let’s not get ahead of ourselves. Before we jump to the possibilities of AI in EHS, let’s focus on taking the steps needed to get there. First, organizations must implement standard EHS processes and data collection across their sites. Then, they need to analyze the data to understand why incidents occur in the first place and how to prevent them. Once this holistic view of risk management is in place, perhaps it is worth revisiting the topic of AI in EHS… but we have a long way to go before we get there.