In PSM-covered facilities, a well-scoped Management of Change (MOC) can make the difference between a safe, efficient modification and a costly, risky one. The Scoping phase—where the change is defined and action items are identified—is the foundation of an MOC’s success. Yet organizations still struggle with incomplete or incorrect scoping due to the complexities of human error, varying levels of experience, and the sheer diversity of possible change scenarios.
The Problem with Incomplete Scoping
For an MOC to be properly executed, it must include a complete and correct set of action items. However, certain types of action items are more prone to being overlooked:
- Inherent items are always included (e.g., PSSR, PHA, approvals).
- Critical Path items are sometimes forgotten early on but are eventually required for the change to proceed.
- Supportive items (like redlining drawings) are often caught late and cause schedule delays.
- Informational items (e.g., updating databases or procedures) are the most missed due to lack of explicit prompts.
Each oversight increases risk, costs time, and can erode compliance.
Scoping Methods: A Comparison
Three common scoping methods are used today:
1. Explicit Scoping
The initiator starts from a blank form and must list all relevant action items. With no structural guidance, this approach leads to high omission rates—only ~51.5% of MOCs are fully scoped under this method.
2. Guided Scoping
Checklists are provided, usually focused on documents and roles (e.g., “Update P&ID?”, “Select reviewers”). This reduces the error rate but still requires significant knowledge from the initiator. The success rate rises to ~80.7%.
3. Asset-Based Scoping
Checklists are framed in terms of plant assets (e.g., “Are you adding safety-critical instrumentation?”). Based on answers, predefined rulesets trigger appropriate action items. This method is far more intuitive and effective, resulting in ~97.6% of MOCs being properly scoped.
Enter the AI Agent: The Next Evolution in MOC Scoping
While asset-based scoping provides structure, it still relies on the completeness and currency of the rule set and assumes the initiator can correctly interpret the asset checklist.
At the 2025 AIChE Global Congress on Process Safety, Dr. Rainer Hoff presented “AI-PSM: Progressing but Not ‘There’ Yet,” where he outlined the promise—and challenges—of integrating AI into Process Safety Management. One promising application is improving asset-based MOC scoping using AI Agents.
How an AI Agent Can Improve Asset-Based Scoping
- Contextual Understanding of Changes
An AI Agent can analyze the technical basis, affected area, and change description to pre-select applicable asset categories and flag ambiguous cases for review. - Dynamic Ruleset Generation
Instead of relying solely on static rulesets, an AI Agent can leverage historical MOC data to suggest action items based on similar prior changes—even for novel or edge cases. - Identifying Gaps and Overlooked Items
By running a probabilistic comparison against a library of past MOCs, the agent can detect patterns of omission and highlight supportive or informational items that are commonly missed. - Tailored Checklists Based on Change Type
The agent can adjust the phrasing, ordering, or priority of scoping questions depending on whether the change is mechanical, procedural, organizational, or experimental. - Continuous Learning
As the AI Agent interacts with more MOCs, its ability to predict needed action items improves, turning asset-based scoping into a self-optimizing system.
Caution: We’re Not “There” Yet
Dr. Hoff cautions that while the foundational models exist, industrial AI in PSM is still maturing. Current limitations include:
- Trust and transparency in AI decision-making
- Data governance and security concerns
- The challenge of distinguishing routine (RIK) from non-routine (MOC-worthy) changes
- A need for explainability and human oversight
Still, AI-enhanced scoping represents the next frontier in reducing MOC scoping risk, particularly as regulatory expectations increase and workforce experience levels shift.
Conclusion: The Future of Scoping
The Scoping phase defines the success of an MOC. While explicit and guided approaches have inherent limitations, asset-based scoping, supported by structured rulesets, has proven to be the most effective method to date.
But as we look ahead, AI-enhanced asset-based scoping holds enormous promise. It offers a way to not only reduce omissions and improve safety but also learn from experience and adapt in real-time—bringing the vision of zero-miss MOC scoping closer to reality.
Want to see how asset-based scoping and AI can work together in your facility? Reach out to Gateway Consulting Group to learn more about FACILEX® PSM solutions, built with safety and intelligence in mind.
