Global Mining Guidelines Group

ARTIFICIAL INTELLIGENCE WORKING GROUP

Artificial intelligence (AI) has great potential to make mining operations safer and more productive. This working group aims to identify the current challenges, educate the industry, keep up with the rapid pace of change in this field, and define collaborative solutions.

ABOUT THE WORKING GROUP

The AI working group aims to facilitate a greater understanding of AI and its applications in mining to enable companies to adopt AI solutions safely and effectively. It also aims to identify the current challenges within the industry, challenges associated with applying AI in mining operations and, define long-term collaborative and innovative solutions in a rapidly advancing technological landscape.

The group is a global network of subject matter experts, operators, suppliers, leaders from inside and outside the mining industry, and those interested in learning more about applying AI in their operations.

OBJECTIVES

  • Develop high-quality material about AI and related emerging technologies for their safe, effective, and efficient application in the mining industry
  • Facilitate a global network of AI experts to share knowledge, best practices, and practical use cases and case studies and examples
  • Enable valuable and open conversations about solving key challenges and advance collaboration about the benefits and challenges associated with adopting AI solutions in mining operations to enable informed decision-making around AI adoption
  • Define AI in a clear and accessible way and support its implementation and long-term sustainability
  • Educate the industry on concepts related to AI

Industry Need

The mining industry is increasingly using AI as a tool to optimize processes, enhance decision-making, derive value from data, and improve safety. Nevertheless, the transition to an AI-enabled mine will look different for every organization, and an inadequate, poorly prepared, or inefficient implementation can result in:

  • Revenue loss
  • Accidental system failures
  • Resistance to change
  • Potential hazards and labour risks

While AI has many benefits for the industry, it is still very new and there is a lot of work that needs to go into successfully applying it. Therefore, it requires a robust foundation of planning, research, communication, and change management.

2023 Focuses

Having industry alignment and education that is widely accepted and understood by all stakeholders of the technology (from OEMs to operators) is key to understanding and enabling the full benefits of AI. Additionally, for industry-wide change, the realization that it won’t be one single technology making the difference, but a series of foundational technologies that will be implemented, magnifies the challenge.

Industry education and alignment around data is crucial to the success of AI technologies because these technologies are data reliant; therefore, industry-wide standardization (on the OEM side) is a requirement.

Additionally, alignment on cybersecurity is needed because it should be viewed from the same lens as safety, not from a competitive aspect, because whenever a new AI technology is implemented, there are risks to the operation.

Increasing industry education about where exactly AI can be beneficial in the mine is needed. This can also be a component to helping with industry alignment by sharing the knowledge of what and where the technology can fit (e.g., ore processing, smart ventilation).

We will deliver this workstream in support with the following focus areas:

  • How can existing AI tools be used to solve common problems (e.g., data standardization)?
  • Roadmap – define end state, start with current actions.
  • As questions get more complicated when determining how a technology can benefit an operation, it can be beneficial to know what emerging technologies exist that can allow a team of humans to ask these questions and to be supported by tools for the questions that increase with complexity. A guideline that can help should:
    • Be flexible and resilient enough to allow growth and integration.
    • Consider cost vs value.
    • Include alignment on adoption to better understand the value of what the cost is, where it can be used, and what it can be used for.
    • Consider the level of maturity required to adopt those types of implementations.

There is a global shortage of AI specialists needed who can use and operate the equipment. SMEs that specialize in one aspect don’t overlap with SMEs in AI, causing a gap in the AI skills required.

Some companies are reluctant to use AI technology (both corporate and members of the workforce) because of lack of training, education, or skills. Because AI can be new to operators, if there is insufficient change management and training, the technology might be neglected, or turned off.

There needs to be more willingness to try new things (in this case, technologies) because there is no single solution. When newer workforce comes to the industry, they expect cutting edge technology and availability (i.e., data availability), but they are not always available, so adaptability to different methods is needed.

Training courses, plans, and upskilling programs are all tools that were mentioned to help with this challenge.

We will deliver this workstream in support of the following focus areas:

  • Addressing the challenge of shortage of AI specialists (identifying level of expertise requirements).
  • Create multi-disciplines Working Groups in developing the Models.
  • Create training material to share.
  • Training in International Society of Automation (ISA) and other existing standards.

One of the most important strategy pieces needed to successfully have AI technology implemented into an operation is to get executive buy-in. With the development of a strategic business case outlining the endpoint and goal of the technology, executive buy-in can be met along with strategic planning to ensure readiness.

 

We will deliver this workstream in support of the following focus areas:

  • Business case development to obtain buy-in from corporate to help demonstrate value and ROI.
    • Standardization across the industry (including across sites)
    • Same region working on the same ores etc.
    • Business cases will be different from site to site.
  • Foundational level and being able to build up the capacity are needed before being able to apply AI.
  • Include recent trends and success cases in the Foundation of AI: A Framework for AI in Mining white paper.
  • “How To” guides for smaller companies on how to implement AI successfully.
  • Challenges with the implementation of AI and the need to change mindsets and skills of the workforce.

WORKSHOP OUTCOMES

NEWS

Artificial Intelligence Workshops

More information about future workshops will be shared shortly.

2023-8-9 Artificial Intelligence Working Group Call

PUBLICATIONS

Open Data Sets for AI in Mining

The objective of this guideline is to provide best practices for data sharing based on existing initiatives for those within the mining industry so they can benefit from the opportunities of open data. This guideline was published in April 2022 and future work on this topic is being discussed by the Working Group. Read the guideline here

Guideline For Sharing Open Data Sets

Foundations of AI: A Framework for AI in Mining

This white paper offers an overview of the process of planning for and implementing artificial intelligence (AI) solutions for mining companies. It addresses a variety of concerns, such as the challenge of establishing data infrastructure, apprehensions about the effect on the workforce and worries about failure after investing substantial time and funds into an AI project. Read the white paper here

ACTIVE PROJECTS

The Artificial Intelligence Steering Committee reviewed the top priorities noted by GMG members throughout workshops held in 2022 and is currently working to determine upcoming projects that can benefit the industry and address the priority challenges.

OTHER ONGOING ACTIVITY

The working group is currently collecting user stories related to the implementation of AI from across the mining industry to create a case study hub. This hub will be a vital resource that can be used by different companies eager to learn from others who have attempted a similar project and can help provide information on what some of their roadblocks, outcomes, and successes from the undertaking looked like.

Do you have a case study you would like to share? Learn more on our case study webpage here

ARTIFICIAL INTELLIGENCE CASE STUDIES

Case Study: Monitoring the Heartbeat of Mining and Metals Production

KNOWLEDGE SHARING PRESENTATIONS

Foundations of Artificial Intelligence in Mining (Video)

The mining industry is increasingly using artificial intelligence (AI) as a tool to optimize processes, enhance decision-making, derive value from data, and improve safety. The white paper aims to build a foundation for mining companies that are planning and implementing AI solutions. Click here to watch.

From idea to product: The application of AI algorithms in mining (Video)

With the mining sector embracing more and more the principles of the industry 4.0, we expect artificial intelligence to be the next revolution in mining. But in data science, between ideas and products there is a long and tortuous road. This presentation will describe how Newtrax has managed to move from ideas to reality for the application of AI technology in mining. Best practices and common traps to avoid will be discussed as well. Click here to watch.

Developing Open Data Sets for AI in the Mining Sector (Video)

The Global Mining Guideline’s AI in Mining Working Group has begun to work collaboratively with its stakeholders to build the case for open data sets specific to the mining sector for AI development. GMG believes this will enhance the ability to build meaningful solutions for the industry by providing typical data relating to assets or operations for training and testing of models, and allowing all parties to have the ability to benchmark solutions and research more effectively. Click here to watch.

Learnings from Staging Petabytes of Data for Analysis in AWS (Video)

AWS hosts a variety of public datasets that anyone can access for free. Previously, large datasets such as satellite imagery or genomic data have required hours or days to locate, download, customize, and analyze. When data is made publicly available in the cloud, anyone can analyze any volume of data without needing to download or store it themselves. We will look at patterns and anti-patterns to watch out for when looking to work with open data across any domain. Click here to watch.

Implementation of Artificial Intelligence in Mining (Video)

The GMG Artificial Intelligence Working Group is currently developing an implementation of AI guideline. Its aim is to leverage lessons learned and case study examples from experience applying AI in mining to help overcome the challenges and barriers to entry and to improve adoption of AI, and as a result, boost efficiency in the mining value chain. Click here to watch.

Open Data Sets for Artificial Intelligence in Mining (Video)

The Industry participants are currently developing a guideline for the collection, cleaning, labelling, and curating of open data sets to help the industry test and train their models for a variety of AI applications.  This guideline is the first phase of a broader project to build open data sets relevant to the mining sector. Click here to watch.

Uncertainty Management Strategies for Industrial Applications (Video)

Danielle Zyngier, Hatch, presented about uncertainty management strategies for industrial applications. We typically talk about AI focusing on data-driven technologies, but AI isn’t only about data-driven methods. There is a lot to be said about AI based on fundamental models and optimizing them, and that can be combined. What is decision-making under uncertainty? Literature talks about many different methods, but they are used for different things.  Click here to watch

Value and Opportunities of IoT and Location Interoperability (Video)

Data and AI are better when they are together. Unfortunately most of today’s Internet of Things (IoT), including location tracking, sensing, and industrial control systems, are disparate systems designed and deployed for one-off applications. That means it is very challenging to aggregate and correlate the heterogeneous data from silo-ed IoT systems for real-time situational awareness and early warnings. This webinar will present the geospatial and IoT standards from the Open Geospatial Consortium (OGC). Click here to watch.

X