AI adoption aims to lift mine safety

The inclusion of generative artificial intelligence in the mining industry is expected to be among the leading business trends for 2024, according to professional services behemoth Deloitte.

Its annual Tracking The Trends report found that, while AI offered a suite of opportunities for companies, improving physical health and safety was one of the key benefits.

In a separate report published by Deloitte AI Institute in September 2023, titled the Generative AI Dossier, the impacts of AI integration in business were explored more deeply.

The report examined cases across six key areas, one of which was enhancing employee safety.

“Contemplating the energy, resources and industrial sectors’ future, generative AI will likely assume a central role in optimising and mitigating health and safety risks by generating worksite-specific safety training that replicates real-world settings and critical scenarios,” the report stated.

“Generative AI can be used to develop personalised and immersive occupational health and safety training materials that allow trainees to be safely exposed to realistic scenarios and thereby reduce or better respond to real OHS incidents.

“Traditional OHS training may only cover some potential scenarios, and it lacks practical opportunities to apply new skills and knowledge.”

Deloitte Australia partner and generative AI leader David Alonso said he believed there were three dimensions to physical safety in the mining sector.

“What can we do about safety before going to site? What can we do about safety when we are on site? What can we do about safety with no one on site?” Mr Alonso said.

“[The last] is ultimately the safest option, but it comes with other challenges.

“If you think about the first one, what can you do before you go to site? Traditionally this involves a lot of generic training … to be inducted and know the safety protocols.

“With the advancement of AI, we are entering a space, which has been moving in the last few years, but even more accelerated now [with] more immersive training.”

Mr Alonso said generative AI could create more realistic training and educate on situational awareness through the simulation of scenarios using technology, such as augmented reality or virtual reality.

“How do you train your people in a much more immersive way before they go to site, so they can really improve their critical thinking, their muscle memory, even?” he said.

“Similar to other industries, you can talk about hyper personalisation.

“That’s personalising training to the individual, and personalising training to the mine site because every mine site is different … different locations, context, environment, dynamic.”

Artificial intelligence can improve worker safety on mine sites.

Deloitte’s Generative AI Dossier also named simulated experiences and personalisation as two key benefits of AI-generated safety training.

VR has capacity to develop environments that replicate operational conditions and simulate OHS incidents, although care and oversight will be needed.

“With realistic scenarios … trainees can navigate hazardous situations, identify risks, and improve their OHS awareness and response capabilities in a safe setting,” the report stated.

“Real-life emergencies can be highly stressful and traumatic.

“Replicating these scenarios virtually could imperil the psychological safety of trainees, and the final design of simulations should be reviewed by human trainers to remove potentially harmful visualisations.”

The report also found customisation of training content was achievable with AI and resulted in safer standards of training.

“Generative AI can be used to customise training materials based on specific job roles, site conditions, or regulatory requirements,” the report stated.

“This technology can analyse large volumes of data, such as incident reports, OHS guidelines or compliance standards, and generate tailored content, including videos, interactive modules, or quizzes.”

Another aspect favouring the adoption of AI for improved safety was its capacity to compile latest updates on current workplace conditions for workers on site.

Mr Alonso said historical data and behaviours could be utilised to predict potential safety issues or identify hazards before an incident occurred.

“If you go to site, every morning there’s a pre-start meeting,” he said.

“Your manager … reviews safety protocols and all the things you need to review before you get to work.

“How do you equip those people with the latest information … on that specific site and specific scenarios about the day? How do you have the most accurate data every morning before those meetings?

“One way is biosensors and biometrics that can track employees from a fatigue standpoint.

“If you have mining organisations equipping their people with wearables to track heart rate, heat, blood pressure, or oxygen levels … then you can correlate this with your fly-in, fly-out roster; how many days they’ve been on site, the type of work they’ve been doing, and then you can ask ‘have we historically identified incidents in similar scenarios?’”

Often companies use AI to improve activity productivity, which Mr Alonso said had a dual benefit for worker safety.

“Say, if you improve your train-loading process with AI to avoid spillage, as an example,” he said.

“If there is no spillage, you don’t need people to go and clean, and therefore you don’t put them in a position where there are risks.

“There are multiple scenarios like that when you are on site.

“Many mining organisations have remote control rooms or systems … autonomous trucks, or drones to detect things and therefore not send people for surveying, identifying corrosion, or whatever else.

“As opposed to sending people, you send machines, or you have machines on site that can do part of the job or some of the tasks that would’ve been done by humans otherwise, and therefore humans are not exposed to those risks.”

David Alonso. Photo: Michael O'Brien

Despite growing implementation of AI in workplaces, Mr Alonso said humans would always remain involved in the mining industry.

“While we are saying all this, it’s also important to think about humans using AI, as there needs to be a way to close the loop, to fit [people] back into the system,” he said.

“As we use AI, and you’re working in a different way, how do you provide feedback so you can improve work processes?

“We give a lot of data, a lot of decisions, options, real time to humans. They largely still make the decisions on a number of elements for the foreseeable future. Or if machines make decisions, humans can still intervene.

“You may eventually upgrade systems ... typical culture in those mining organisations is around continuous improvement.

“Continuous improvement doesn’t stop with the use of AI. It’s just adding another dimension.

“The human worker is still a critical part of mining operations. Yes, tasks may be replaced by machines, but other tasks will be created.”

These ‘other’ jobs are also likely to be safer and enable more productivity as operations evolve with the assistance of AI.

Mr Alonso said physical safety would always be the top priority for mining organisations, and collaboration across industry would further enhance the improvement of this.

“Ultimately, all workers in the regions of Western Australia are important for the industry,” he said.

“Therefore, how are we going to use AI across industry to perhaps organise training, certifications, and multi-party workflows?

“That brings another lens to how we use AI, not only for single organisations but for the workforce across the industry, and I think that can also add a lot of value for all the workers, but also for all the organisations.

“There are many positive aspects of using AI for mining, and specifically focused on safety so everyone goes home safe after their shift.”