Safety Compass: How AI is Changing the DNA of HSE

Artificial Intelligence as a New Tool for the HSE Specialist

The digitalization of HSE processes is moving to a new level. While automation was previously limited to transferring paperwork into Excel and specialized software, generative neural networks and artificial intelligence (AI) systems are coming to the forefront today. The presentation details practical experience in using AI tools that help specialists eliminate routine tasks and free up time for analytical work and live communication with employees.

The speaker explains how generative models (such as chatbots) can create content, analyze data, and even write code based on text prompts. It is important to understand that AI is not a replacement for a specialist, but rather a "digital intern" that needs to be trained on specific examples and corrected when mistakes are made.

The "Health Compass" Case: Automating Epidemiological Risk Control

One of the key projects presented in the webinar is the "Health Compass" system, developed to manage risks during employee business trips to different countries. Previously, vaccination control and informing about the epidemiological situation were done manually, which required enormous resources as the number of trips increased.

What the system does:

• Analyzes data in real time: collects information from WHO and CDC Travel Health websites about disease outbreaks and recommendations.
• Categorizes countries by risk level: from low (basic hygiene advice) to high (mandatory vaccination and symptom monitoring after the trip).
• Personalizes recommendations: takes into account the employee's route and vaccination history, providing specific instructions before departure.

The speaker demonstrates by example how a prototype of this system was built in just a few hours using the low-code platform Replit. This made it possible to quickly test the idea, create a clear technical specification, and successfully defend the project before the internal information security service for subsequent transfer to the company's secure servers.

Low-code Platforms and Incident Forecasting

The presentation also touches upon the use of Replit and Bolt platforms to create interactive dashboards and incident registration systems without deep programming knowledge. The speaker demonstrates how, with a properly crafted prompt, a working application prototype can be obtained in a matter of minutes.

This approach allows you to:

• Visualize data: create incident heat maps, days-without-injury counters, and safety ratings.
• Automate categorization: a bot can analyze a text description of an incident (e.g., "driver fell asleep at the wheel") and automatically determine the incident type, severity, and cause (human factor).
• Improve interaction with IT: having a working prototype greatly accelerates the process of developing a full-fledged product by specialized professionals.

Other AI Application Scenarios in HSE

In addition to large projects, the speaker shares examples of everyday AI use to optimize routine tasks:

• Risk assessment: generating risk assessment maps from workplace photos with subsequent refinement by a specialist.
• Incident investigation: analyzing protocols with an AI model to get a "third opinion" and identify non-obvious factors.
• Training and briefings: creating scenarios, instructions, and visual materials, which significantly increases employee engagement.
• Feedback analysis: summarizing employee opinions to develop plans for improving safety culture.

What you will learn from this webinar:

• How to start using generative neural networks in the daily work of an HSE specialist?
• How to create a risk management system prototype in a few hours using low-code platforms?
• How to automate the collection and analysis of epidemiological data for traveling employees?
• How AI can help categorize incidents and create risk assessment maps?
• How to properly formulate requests (prompts) to get high-quality results from neural networks?