The Future of HSE Through the Lens of AI Development Trends

Artificial intelligence is rapidly transitioning from a category of technological hype to a real working tool. Against the backdrop of growing information overload, talent shortages, and increasing business process speeds, digitalization is becoming a natural and inevitable stage of industry development. Attention to these technologies is confirmed at the global level: the International Labour Organization (ILO) dedicates thematic days to digitalization, and national-level strategies for continuous learning are being formed. Large businesses are already introducing AI implementation KPIs for top management. In his presentation, Rinat Fatkhutdinov examines in detail how to adapt occupational health and safety processes to new realities while avoiding typical automation traps.

The Illusion of the Magic Button: Why 95% of AI Projects Fail to Deliver Results

Despite the fact that 97% of large companies claim to be implementing or planning AI initiatives, only half see real value in them. Drawing on research from the Skolkovo Business School, the speaker notes that the vast majority of corporate artificial intelligence projects end at the pilot stage or fail to produce business results.

The problem lies not in the technologies themselves, but in fundamental process errors:

• Lack of a data culture. Unlike fintech, where information is the primary product, industrial companies often lack a structured system for collecting and storing high-quality data. Without a standardized database, algorithms have nothing to learn from.
• Automation for the sake of automation. Attempts to implement algorithms across all processes without prior audit and ranking lead to the digitization of chaos.
• The gap between IT and production. The absence of process owners and clear metrics from specialized departments leads to the creation of unviable products detached from reality.

The Role of "Knowledge Brokers" in HSE Digitalization

IT department specialists do not know the daily "pain points" of safety engineers. For successful technology implementation, companies critically need "knowledge brokers"—proactive ambassadors within the HSE departments themselves. These are specialists who deeply understand the specifics of production processes, are ready to study new tools, and can translate business requirements into a language understandable to developers. They are the ones who help overcome the barrier of resistance to innovation on the ground and connect technology to real tasks.

Practical Cases: From Failures to Labor Cost Reduction

The presentation analyzes a series of contrasting examples from corporate practice. Unsuccessful scenarios are always linked to ignoring the real needs of users. For example, when an oil company implements a corporate model that performs tasks worse than available neural networks (such as DeepSeek), employees simply refuse to use it. Another extreme is the development of IT solutions in a vacuum, where production staff are not even aware of the existence of a product created for them.

Successful cases are built on engagement and phased scaling. In one mining company, a secure internal analog of ChatGPT was created. The tool was integrated with corporate accounting systems and Robotic Process Automation (RPA) technologies. The result was a 20% reduction in labor costs for routine intellectual tasks.

The main management takeaway lies in following a three-stage model: first, mass training of personnel and removing fears of technology, then implementing and accurately evaluating one successful pilot project, and only after that, scaling the successful experience to the entire company.

What You Will Learn from This Presentation

• Why do most corporate AI implementation projects fail, and how can you avoid becoming part of that statistic?
• Who are "knowledge brokers," and why is effective HSE digitalization impossible without them?
• How to safely use generative neural networks within the corporate perimeter of mining and industrial companies?
• What three steps must you take to begin integrating artificial intelligence into HSE department operations to achieve measurable results?