The November 2024 DROPS Wells Forum showcased groundbreaking advancements in AI Solutions for DROPS Prevention in the drilling sector. A standout presentation by KCA Deutag focused on their FX-RZS (Red Zone Supervision) system, an AI-supported solution developed with Shell Oman to enhance operational safety in restricted zones. The forum also introduced the concept of “DROPS GPT,” a specialized ChatGPT model designed to transform access to the DROPS Recommended Practices, offering unparalleled support for knowledge sharing, training, and decision-making in the field of dropped object prevention.

KCA Deutag AI Supported Red Zone Monitoring

The meeting presented an in-depth overview of the FX-RZS (Red Zone Supervision) system, which leverages artificial intelligence to enhance safety by monitoring and managing hazardous zones in drilling environments. The system’s development, technical details, operational impact, and future plans were discussed.

Holger Egbers introduced FX-RZS as a groundbreaking solution for safety in onshore and offshore drilling operations. The system uses AI-powered object detection to monitor Red Zones, dynamically adjusting boundaries based on operational data such as equipment activity or pressure levels. Key features include real-time alerts via a traffic-light interface and integration with existing rig systems, offering a significant layer of safety and control. The interface communicates safety conditions clearly: steady red for active hazards, flashing lights for personnel detection, and green for safe conditions.

Egbers emphasized the system’s performance improvements over time. Initial implementation recorded breach times averaging 48 minutes per hour, which decreased to 26 minutes after enhancements. This reduction was achieved through better analytics, adjustments to the interface, and increased operator awareness. The presentation demonstrated how detailed data analytics identify trends in zone breaches and guide process improvements to mitigate risks.

Challenges in implementing FX-RZS were highlighted, such as the need for continuous training of AI algorithms to recognize various objects and manage complex scenarios. Despite these challenges, the system adds a critical safety layer by reducing personnel exposure to hazards, complementing the Shell Dropped Object Prevention Guide.

Future developments include advanced real-time data analytics for predictive warnings, expanded system integration for additional operations, and refined manual override capabilities. Attendees recognized the potential of FX-RZS to revolutionize safety management in drilling environments and agreed on the need for ongoing collaboration to refine and expand its functionalities.

Yahya Al Rubaiai from Shell Oman shared his perspectives, having monitored the implementation in the field.

download the presentation here  KCAD_FXC_RZS_DROPS Forum_112024

DROPS GPT

In this Forum, we also explored the possibility of having a ChatGPT specifically trained on the DROPS Recommended Practices. Such a “DROPS GPT”  could offer numerous benefits to members of the DROPS Asia Chapter and the broader community focused on dropped object prevention. Here are some of the possible advantages:

1. Instant Access to Expert Knowledge

• Members can ask questions about best practices, policies, and recommendations related to dropped object prevention and receive accurate, context-specific responses.
• Eliminates the need to manually search through documents or training materials, saving time and improving accessibility.

2. Consistency in Information

• Ensures that all members and stakeholders are aligned with the latest DROPS guidelines and recommendations.
• Reduces the risk of misinterpretation or miscommunication about safety procedures or requirements.

3. Enhanced Training Support

• Facilitators can use the ChatGPT model during training sessions to provide real-time answers and clarification, improving the learning experience.
• Interactive engagement with the model can help reinforce knowledge retention among trainees.

4. Scalability and Accessibility

• Can support a large number of users simultaneously, making it ideal for global operations where members may be in different time zones.
• Provides 24/7 availability, ensuring that members can access assistance whenever needed, even in remote locations.

5. Improved Decision-Making

• Helps members make informed decisions quickly by providing guidance based on established practices and evidence.
• Reduces the time spent deliberating or consulting multiple sources, particularly during critical operations.

6. Support for Multilingual Users

• If configured to support multiple languages, the model can help non-native English speakers understand and implement DROPS guidelines more effectively.

7. Customization for Specific Needs

• The model can address unique challenges faced by different companies or projects within the DROPS community.
• Members can get tailored advice for specific scenarios, such as inspections, risk assessments, or tool selection.

8. Knowledge Retention and Updates

• Serves as a centralized repository for DROPS knowledge, preserving expertise even as personnel change within the organization.
• Can be updated with new practices or insights as the industry evolves, ensuring ongoing relevance.

9. Cost-Effective Knowledge Sharing

• Reduces the need for extensive manual support, as the model can answer routine queries, freeing up human resources for more strategic tasks.
• Minimizes travel and training costs by providing a virtual assistant for knowledge dissemination.

10. Integration with Other Tools

• Can be integrated into existing platforms, such as the DROPS Metaverse or other training and safety management systems, to enhance user experiences.
• Provides an additional layer of interactivity in virtual and in-person environments.

In summary, a ChatGPT trained on DROPS Recommended Practices would serve as a powerful tool to enhance the accessibility, understanding, and application of critical safety guidelines, ultimately contributing to a safer work environment and more efficient operations. If you could like to be involved in a workgroup to develop this Chat GPT, please contact the DROPS Asia Chapter Secretary Joachim van der Meulen