AI in the EHS industry
New technologies and the ways of working that they bring about, present environmental, health and safety (EHS) managers with opportunities to better manage risk and enhance compliance and operational performance. One of the most talked about innovations in recent years is artificial intelligence (AI).
Still in its infancy, AI has the potential to be truly transformative in the EHS space and has been used in worker management systems for almost a decade to generate data on workspaces, workers and the work they do (https://osha.europa.eu/en/publications/artificial-intelligence-worker-management-existing-and-future-regulations). However, is there a risk that as AI becomes more prevalent and sophisticated, its applications could make EHS managers redundant?
Most observers in EHS software and in frontline operations seriously question that this potential scenario could become a reality. In fact, the general consensus is that EHS managers will maintain operational oversight of AI applications, not least to ensure the data generated is accurate and remains fit for purpose.
Even so, what safeguards do businesses and organisations need to put in place to ensure that the fictional HAL 9000 AI character from Arthur C Clarke’s Space Odyssey doesn’t become a reality, especially in high-risk industries?
There is a much coined-phrase in safety industry circles which goes “If you think safety is expensive… try having an accident”
The phrase refers to an attitude that can be prevalent in less mature businesses that investing in safety as a preventative action is far too costly. However, the reality is that the financial, not to mention human and reputational, cost is even greater if an accident does occur. Mature businesses are all too aware of the negative publicity that surrounds serious incidents and the impact on brand reputation and the bottom line.
With this in mind, it is perhaps inconceivable that businesses that are in a position to benefit from AI would risk using the technology for safety critical decisions without humans determining how the technology is applied, monitored and acted on.
This thinking is reflected in the European Agency for Safety and Health at Work’s AI policy brief, which notes how a regulatory framework will be needed to mitigate any negative effects of AI-based worker management systems and makes recommendations to resolve any issues where regulatory gaps exist.
These would include making sure that humans always remain in control of AI systems and establishing a clear line of responsibility, including oversight mechanisms.
In reality, it is probably safe to assume that AI systems will only be used in scenarios where low-risk decisions are made; typically ones that affect EHS managers’ routine and manual tasks.
Although AI could make automated or semi-automated decisions, for instance assigning employees to undertake audits, allocating resources or monitoring activities, the EHS manager will still be calling the shots.
To give an example, Daten & Wissen, a Mumbai-based IT services and consulting firm, suggests that AI is perfect for video analytics.
How does AI help EHS managers?
To start with, AI can gather data pulled from surveillance cameras quickly to alert EHS managers to potential risks so preventative action can be taken. For instance, AI could monitor activities to see whether workers entering a hazardous area are wearing personal protective equipment (PPE) such as hard hats and face masks. Likewise, AI could be applied to refuse unauthorised access to restricted areas.
EHS managers will still need to weigh up the benefits and costs of introducing new technologies like AI into their day-to-day operations.
A reduction in overall business costs is likely to be a significant factor in whether AI is applied. As Daten & Wissen note, AI driven video surveillance and analytics in a warehouse and/or factory could save businesses the cost of employing security guards.
There is also the time-saving element. Because video analytics performs monitoring in real-time, when something suspicious or out of the ordinary is spotted, alerts can be raised within seconds prompting an immediate response.
Arguably, there is also an improvement in risk management because AI can monitor worker safety more accurately.
What this means in practice is that EHS managers could invest precious resource and time in more strategic considerations and focus their efforts on preventative actions to minimise the risk of incidents escalating or happening in the first place.
In this respect, the critical data that AI is able to process and collate far more effectively and quickly than a human can help to inform strategy. With the EHS managers overseeing the AI operations, it could, for example, be possible to improve intelligence around lagging and leading indicators and identify any important trends.
Applications like those described above could be a potential game-changer as long as they are managed correctly and the right data is pulled out and acted on. And there will always be limitations in terms of what AI can do.
For example, although the technology might be able to identify if someone is wearing PPE, can it determine whether the PPE is being worn correctly and if the person who is working with chemicals removes the PPE correctly in the decontamination area to avoid cross-contamination?
AI can add great value to EHS management systems. However, it is clear that the EHS manager’s future is safe and secure and that this fast developing technology has the potential to significantly enhance this critical role.
Finally, it is also worth noting that there will always be a need for EHS managers to support workers who are dealing with AI systems, providing much needed compassion and empathy when there are difficult situations to manage. To learn more about how Protex AI is using camera software to help EHS safety managers embrace a proactive safety culture, chat to one of our product experts here 👈🏼