AI in Mining panel: Lessons learned on readiness, change management and strategies

By Francine Harris

A group of experts representing both mining companies and technology suppliers discussed practical approaches to AI in mining on a panel at the GMG-CIM AI in Mining Forum on May 1 in Montreal. The panellists: Michel Plourde, Director Integrated Operations – Innovation & Technology at ArcelorMittal; Mark O’Brien, Manager – Digital Transformation at CITIC Pacific Mining; Dave Shook, President and CEO of ShookIOT; Kevin Urbanski, NextAI Entrepreneur and co-founder of Rithmik Solutions; and Michel Dubois, Vice-President QA & Artificial Intelligence at Newtrax, were guided through a robust discussion by moderator Andy Chapman, Technical Manager, Mining Solutions at Peck Tech.

AI readiness

Urbanski listed six elements of an AI-ready mine: someone internal who knows what AI can do; the problems to solve; the data to create models to solve those problems; the agility and the culture to pull it off; change management; a support plan.

AI-readiness involves people, processes and technology, Shook agreed. You need experts supporting the journey, the infrastructure to sustain AI in the long term and well-organized and accessible data.

Plourde said that the process starts with understanding your advanced analytics capabilities, which can lead to discovering whether you are AI-ready. While ArcelorMittal has had several projects, they are still struggling because they do not have a full dataset available. “Digital deployment capability and maturity within the organization is key before proceeding with first steps in delving in AI,” he added.

Mining companies might not know if they are ready until they try something. Dubois advised to just get started, to “do a very small project to find out [what AI is].” However, O’Brien advised a more spirited approached, suggesting to “jump in and experiment.” For O’Brien, motivation is also a critical piece because AI projects can only be successful with strong buy-in from senior management. In an interview before the panel, he identified two motivation types: “downside” motivations focus on how AI can help solve problems; “upside” motivations focus on how AI can “improve productivity, efficiency, equipment utilization and process optimization to drive better outcomes from our supply chain and operations.”

People and change

When asked about change management, the panellists agreed that education and communication were key components and that safety should be at the forefront of all messaging.

In Plourde’s experience at ArcelorMittal, it’s all about communication and transparency. “You tell everybody everything from the get-go,” he said. “Then you repeat it again and again and I think you hope people will start to understand. There’s no cutting corners, especially for solutions around which people might have privacy concerns, such as fatigue management, extensive change management processes are required.”

Credibility is at the core of change management for Shook. He stressed that it is obvious when it works. For example, he said, “predictive maintenance capabilities earn their credibility.” For people’s privacy concerns and mistrust associated with operational solutions like machine vision, it “comes down to operations management credibility.”

Dubois also suggested changing as little as possible, explaining that Newtrax develops solutions that allow customers to use the systems they already have when they can.

Small wins and broader strategies

One audience member noted that most of the discussion on topics like machine vision and predictive maintenance covered operational problem-solving and wondered if they were thinking of broader strategies.

Plourde responded that in their digital initiative projects they are trying to “chase bottlenecks” and looking at where they can get the most value quickly. In the early stages, it is important to focus on short-term wins and invest in those. Dubois added that short-term solutions make a big difference and can feed into longer-term strategies.

One way that focusing on small wins can lead to broader strategies is how they can help with data management. “One of the great things about [AI] is that it’s really making people look at their data,” said Urbanski. In an interview before the panel, he explained that “an early use of AI could be to explore the potential of the data and point to gaps that exist in it.”

The relationship between Industry 4.0 and AI 

One of the final discussions was on the relationship between AI and Industry 4.0. An audience member noted that the term Industry 4.0 isn’t discussed as much anymore and asked whether AI is coming too fast and replacing it.

“AI isn’t really a replacement for Industry 4.0,” said O’Brien. “It is an enabler.” Interconnected digital systems are at the core of Industry 4.0 and AI is a capability that enables them.

The language is what has changed. Plourde explained that when we talk about digital transformation and digital capability, we are talking about Industry 4.0. Dubois suggested that the industry is talking less about Industry 4.0 because it is “starting to be taken for granted.”

Shook added that while AI is moving faster than he has ever seen, it is just part of a great transition happening in industrial technology. “AI has been around for a long time. It is just now progressing because data management and processing have caught up.”

Key takeaways

Chapman concluded the panel by emphasizing the need for a pragmatic and incremental approach to technologies like AI. Reflecting on the discussion, he says “the adoption of new technologies can be a very slow process and its success is heavily reliant on proper change management spanning the entire workforce.” He added, “we need to ‘walk before we run’, proceed with cautious optimism, and ensure all stakeholders are brought along on the journey.”



X