ML23310A142

2023 CNS Diet Conference - NRC Slides

ML23310A142
Person / Time
Issue date:	11/08/2023
From:	Luis Betancourt NRC/RES/DSA
To:
Betancourt L
References
Download: ML23310A142 (15)

Text

Artificial Intelligence Preparedness:

A Regulatory Perspective Luis Betancourt, P.E U.S. Nuclear Regulatory Commission 2023 CNS DIET Conference International Regulatory Perspective November 8, 2023

Artificial Intelligence Landscape and the NRC 2

ACTIVITIES Wide range of AI meetings, conferences, and activities Industry wants to use AI Federal actions for accelerating the use of AI in government operations AI Strategic Plan to prepare staff to review AI NUCLEAR INDUSTRY (EXTERNAL)

OTHER CONSIDERATIONS AND OPPORTUNITIES (EXTERNAL)

NRC Evidence Building Priority Questions Internal interest in researching AI-based tools ranging from AI-embedded in commercial applications to custom programming INTERNAL TO THE NRC

Nuclear Industry AI Landscape

• Industry Project Categories

• Increasing existing economic efficiency

• Plant condition monitoring

• Process improvement and cost reduction

• Plant automation

• Sensor-enabled degradation assessment

• Example Operating Reactor Applications Areas

• Non-destructive examination

• Advanced remote monitoring

• Corrective action process automation

• Core design optimization

• Generative document preparation

• Physics-informed surrogate models

• Example Advanced Reactor Application Areas

• AI/ML-enabled digital twins

• Autonomous operation of backend processes

• Design optimization 3

4 The AI Strategic Plan consists of five strategic goals:

• Goal 1: Ensure NRC Readiness for Regulatory Decisionmaking

• Goal 2: Establish an Organizational Framework to Review AI Applications

• Goal 3: Strengthen and Expand AI Partnerships

• Goal 4: Cultivate an AI-Proficient Workforce

• Goal 5: Pursue Use Cases to Build an AI Foundation Across the NRC Vision and Outcomes

• Continue to keep pace with technological innovations to ensure the safe and secure use of AI in NRC-regulated activities

• Establish an AI framework and cultivate a skilled workforce to review and evaluate the use of AI in NRC-regulated activities Draft Available at ML22175A206 Final available at ML23132A305 4

Artificial Intelligence Strategic Plan Overview

5 5

Artificial Intelligence Project Plan Overview Available at ML23236A279

• The AI Project Plan describes how the agency will execute the five strategic goals from the AI Strategic Plan

• Provides estimated timelines for various task completions within each Strategic Goal

• AI Regulatory Framework

• AI Governance

• Domestic and International Activities

• Workforce Planning

• AI Ecosystem

AI Research Determine approach to assess AI (e.g., XAI, trustworthiness, etc.)

Framework and Tools Public meetings to inform key activities Clarify the process and procedures for AI regulatory reviews and oversight Consider options for long-range changes for AI regulatory reviews and oversight that may require rulemaking Development of AI standards and identify where gaps exists Communications Agency-wide internal communications and coordination to harmonize AI activities 6

Outcome: Develop an AI framework to review the use of AI in NRC-regulated activities KEEPING THE END IN MIND: DETERMINING THE DEPTH OF REVIEW Goal 1. Ensure NRC Readiness for Regulatory Decisionmaking

Safety Significance AI Autonomy Security Explainability Model Lifecycle Regulated Activity Regulatory Approval Application Maturity AI Characteristics for Regulatory Consideration 7

NRC AI Regulatory Considerations (1/2)

• Existing Guidance - Traditional safety, security, software, and systems engineering practices are still applicable as the starting point for good engineering practice.

• Establishing a Trustworthy System - Explainability exposes a chain of decision-making for potentially complex logic that is easily interpretable by anyone unfamiliar with the AI system design. This applies to all stakeholders which include reviewers (e.g., regulators) as well as system users.

• Safety Principles using Risk or Determinism - In the absence of the ability to quantify risk, there are good engineering principles (e.g., defense-in-depth) that can be used to guard against unintended consequences.

• Open-Source Tools - Use of open-source tools are not precluded, but using non-specialized software solutions means that there are steps taken to rigorously confirm the safety and security of the implemented solution.

8

NRC AI Regulatory Considerations (2/2)

Failure and Consequence Identification - A first step as part of AI systems engineering, a formalized process to quantify the hazards and modes of operation can be considered to ensure adequate system design.

Data Provenance - Based on a graded approach, the modeling data may have a variety of various pedigrees based on the application area (e.g., safety significance).

Model Updating - Models need to be maintained to avoid performance degradation and kept consistent with the pre-determined change control and notification process for that application Human and Organizational Factors - The context of operation needs to consider the handover to human operation, immediacy for human action, or if placement in a safe stable state is required based on the operational context.

Extensive Application Areas - A variety of regulatory requirements apply to various potential AI application areas. Existing requirements may range from evaluation of sufficient functional performance up to specific requirements to ensure AI system safety and security.

9

Strengthening and Expanding AI Partnerships

• Enhancing and leveraging existing Memoranda of Understanding (MOU)

• Participating in federal AI working groups and maintaining awareness of other AI regulatory activities

• Engaging with international counterparts interested in AI for nuclear

• Maintaining continual engagement on state-of-the-art research 10 GAINING VALUABLE INFORMATION TO BENCHMARK AI ACTIVITIES Enhancing and Leveraging MOUs New Data Science and AI Addendum Operating Experience and Data Analytics International Collaboration Maintaining Federal Awareness

Tri-Lateral Collaboration on AI and Innovation

• CNSC/UKONR/USNRC Innovation Hubs established a trilateral relationship in March 2022 to share knowledge and discuss disruptive, innovative and emerging technology (DIET)

• Quickly became evident there was common interest and approaches to AI being used in regulated nuclear activities

• Three regulators agreed to work together to produce and publish a trilateral AI principles paper oExpected publication date: Spring 2024

• Common regulatory approaches each regulator may use regarding AI highlighted 11

Exploring AI At The NRC

• Improving Regulatory Efficiency

• Evidence Building Plan

• ADAMS Content Search

• Investing in the NRCs Future Focused Research Program

• Machine learning to inform inspection planning

• Characterizing cyber security using AI/ML

• Applying NLP to NRC regulatory documents 12

Moving Forward and Stakeholder Engagement

• NRC must remain vigilant AI technologies are entering the nuclear space from both the front and back door

• NRC has been proactively working to understand this evolving technology space to identify technical and regulatory challenges and gaps, gather insights on potential use cases, and develop institutional knowledge

• We are working to ensure we have the staff with the knowledge, skills, and ability to effectively regulate these new technologies

• We are hiring! See job opportunities at: https://www.nrc.gov/about-nrc/employment.html

• Future Activities o Advisory Committee on Reactor Safeguards Subcommittee meeting on AI (November 15, 2023) o Regulatory framework applicability assessment of AI in nuclear applications (Summer 2023-Spring 2024) o USNRC hosting IAEA AI Deployment Technical Meeting at Rockville, MD (March 18-21, 2024) 13

Contact Information Luis Betancourt, PE Chief, Accident Analysis Branch Division of Safety Analysis Office of Nuclear Regulatory Research luis.betancourt@nrc.gov 14 Visit us at https://www.nrc.gov/about-nrc/plans-performance/artificial-intelligence.html

Abbreviations

• ADAMS -Agencywide Documents Access and Management System

• AI - Artificial Intelligence

• CNSC - Canadian Nuclear Safety Commission

• DIET - Disruptive, Innovative and Emerging Technology

• DOE - U.S. Department of Energy

• EPRI - Electric Power Research Institute

• FDA - U.S. Food and Drug Administration

• FY - Fiscal Year

• GSA - U.S. General Services Administration

• IAEA - International Atomic Energy Agency

• ML - Machine Learning

• MOU - Memorandum of Understanding

• NLP - Natural Language Processing

• NRC - U.S. Nuclear Regulatory Commission

• ONR - U.K. Office for Nuclear Regulation

• NIST - National Institute of Standards and Technology

• XAI - Explainable AI 15