Artificial intelligence in risk and safety management

8 minutes07/08/2023

Why should you rely on artificial intelligence (AI) when it comes to occupational safety? Because AI solutions excel in many disciplines that are critical for workplace safety: collecting, organizing, and analyzing substantial amounts of data, providing and processing information from sources such as legal texts, simulating situations, and monitoring processes and facilities.

What exactly is AI? AI refers to the development of computers or machines capable of performing tasks that would normally require human intelligence. As a result, AI is applied in various ways in the field of occupational safety. Click through the directory to access the different areas:

Chatbots in occupational safety

• Summarizing and simplifying legal texts
• Audit support

AI in risk management

• Identification and reduction of accident risks
• Optimized evaluation and prevention
• Automation of repetitive tasks

Predictive maintenance

• Data collection
• Data analysis
• Machine learning and prediction
• Prevention

Video monitoring

AI for training and instructions

Disclaimer

This text was not written by artificial intelligence. AI was only used for targeted search and evaluation of relevant information. Especially in occupational safety, AI solutions should not be at the wheel, but they are welcome in the passenger seat. As co-pilots, they can provide valuable support to those responsible for occupational safety.

Chatbots in occupational safety

Chatbots have undergone significant development in recent times. One impressive example is Chat GPT - an AI application with remarkable language abilities. This artificial intelligence can assist occupational safety professionals by answering specific questions, autonomously creating texts, summarizing expert contributions, and even providing suggestions and ideas for targeted actions.

Caution: The information used to develop Chat GPT only goes up to the year 2021. Therefore, the AI may not be aware of the latest developments in occupational safety. It may provide information that does not correspond to current facts, which the developers refer to as “hallucinations.” It is therefore important to check the responses given by Chat GPT.

Nevertheless, the possibilities that today’s chatbots open up in occupational safety are revolutionary. Artificial intelligence in occupational safety can simplify various tasks and significantly improve process efficiency. The following two examples illustrate this:

Summarizing and simplifying legal texts

In occupational safety, many regulations, directives, and legal texts are not only lengthy, but often difficult to understand. Artificial intelligence chatbots can help solve this problem.

Open the chatbot and type 'Please summarize the following regulation in plain language,' then copy the text of the desired occupational safety regulation into the input field. ChatGPT-4 can process up to 50 pages in this way. The desired result are available in seconds.

You can use the provided content as the basis for employee briefings or presentations. However, fact checking is essential. The EHSQ professional must have previously read the regulation and be familiar with its content. With the ongoing development of AI, especially in conjunction with other applications, the results are expected to become more reliable in the future.

Audit support

Preparing for audits is one of the key tasks of occupational safety professionals, and chatbots can make this work more efficient. For example, they can help find relevant regulations and laws, prepare appropriate training on occupational safety topics or evaluate the results of previous audits.

Enter the following text in Chat-GPT: “Create a checklist of all necessary steps for an internal fire protection audit.” The AI will immediately provide the top ten items to consider.

With AI-generated checklists like these, you can check whether all the necessary steps have been considered and all required documents have been prepared. In short, the chatbot helps you comply with occupational safety regulations and avoid pitfalls.

Other uses for chatbots in occupational safety:

• Researching best practices for specific safety topics
• Providing information on new or changed regulations
• Creating safety plans
• Modifying procedural instructions
• Automating the response to safety-related questions via email or chat
• Analyzing data from the incident management database
• Automatically notifying of critical safety conditions
• Developing occupational safety training materials
• Researching ideas for practical exercises on specific safety topics
• Analyzing images and converting images to text

AI in risk management

The applications of artificial intelligence in risk management are many and varied. AI applications can monitor critical infrastructure, perform automated facility inspections or be integrated into the event database. The latter provides measurable risk management benefits to professionals and executives.

Identification and reduction of accident risks

Integrated with the incident database, AI helps you analyze all recorded accidents, near misses, and unsafe situations. With the help of artificial intelligence, you can quickly and comprehensively identify potential hazards and effectively reduce accident risks.

Optimized evaluation and prevention

AI in risk management shows no signs of fatigue. It can evaluate massive amounts of data in a short amount of time, flawlessly. AI applications identify patterns, correlations, and trends, and present the data in a structured and clear manner for human colleagues. And that's not all: Based on the analysis, artificial intelligence can make predictions and suggest targeted actions and preventive measures derived from the data.

Automation of repetitive tasks

In addition to its speed of data evaluation, AI-based systems can automate repetitive tasks, freeing up human resources. This allows them to save time and focus on more important tasks. Some of the repetitive tasks that AI can handle in risk management include:

• Continuous analysis
• Regular data monitoring
• Cyclical reporting
• Answering questions about how to use the event database

Tip

AI works with the data you provide. First, make sure that your incident management, including reporting of accidents and near misses, and your set of key performance indicators are solid. Two critical prerequisites for encouraging employees to report incidents are 1. a strong failure and feedback culture, and 2. simple digital solutions that support quick and user-friendly reporting. Veltec GmbH & CoKG, one of Europe's leading specialists in technical services for the process and power plant industry, increased its near miss reporting by 70 percent by implementing EHSQ software. Read more: https://www.quentic.com/success-stories/veltec/.

Predictive maintenance

Predictive maintenance refers to the proactive maintenance or repair of a machine before it fails unexpectedly. Artificial intelligence opens up a wealth of new possibilities for occupational safety, for example, in power plants or with machines and equipment that handle toxic substances, high electrical voltages, and potentially explosive atmospheres. Artificial intelligence in predictive maintenance eliminates the need for periodic maintenance on such assets, not only saving costs but also reducing the frequency with which employees are exposed to the risks of hazardous work:

Data collection

AI sensors on machines collect various measurements that reflect the condition of the machine. This can include temperatures, pressures, vibrations, noise levels, or flow rates. Additionally, the AI is fed with historical data resulting from malfunctions or damage.

Data analysis

Artificial intelligence analyzes available data in real time, identifying correlations and relevant patterns essential for predictive maintenance.

Machine learning and prediction

Based on the analyzed data, the AI application "learns" to make predictions and issue warnings about impending disruptions and failures.

Prevention

Before something happens, humans can proactively intervene. Artificial intelligence in predictive maintenance not only minimizes machine downtime, but also helps improve safety. AI can detect damage well in advance that could lead to dangerous malfunctions or other risks.

Video monitoring

Today, camera systems can be cost-effectively integrated into almost any process, including occupational safety. Artificial intelligence is an ideal intelligent extension for video surveillance. For example, AI applications can keep an eye on reactors, pumps, gas pipelines, or high-voltage motors.

However, these examples only scratch the surface, as the industry trend is moving towards autonomous robotics and transport systems. The use of AI is recommended to identify and track these systems in real time: Are the autonomous helpers following their planned routes? Are there any unusual movements or operations? Are the materials being transported being stored safely? Artificial intelligence evaluates the situation and alerts in case of  deviations – long before an accident can occur.

AI for training and instructions

Artificial intelligence in occupational safety can also help professionals prepare training and instructions. Here's how it works: Humans feed selected data to the AI, such as their company’s best practices, standardized procedures, or guidelines, risk assessments, and industry-specific legal requirements.

Based on this content, current AI applications are able to plan and support the creation of training sessions or instructions. In addition, artificial intelligence in occupational safety can be a significant help even after the training has taken place. It can analyze participant performance, provide feedback, and make targeted adjustments to the training based on the results.

Conclusion

Artificial intelligence is a powerful tool in occupational safety, with significant potential in areas such as data and text analysis, risk management, predictive maintenance, video monitoring, and training. However, while AI can identify patterns and correlations, it lacks a true understanding of safety. Human expertise and experience remain essential.

In the future, the vast amount of content generated by chatbots will be reintegrated into the World Wide Web. As a result, AI will increasingly rely on data it has generated itself. However, innovative ideas, original knowledge, and creative approaches will still originate from humans.

Much like modern occupational safety software, AI greatly improves efficiency, but it cannot replace the human element. Its real value lies in saving time and effort for everyone involved. In this way, AI enables occupational safety professionals to focus their time on safety leadership rather than administrative or repetitive tasks.