Introduction
It has been eight months since I stood down as the chairman of the IET’s Project Controls Technical Network after serving on the steering committee for many years. This decision was driven, at least in part, by an increasing desire to advance the body of knowledge relating to the adoption of artificial intelligence in the field of quality, assurance, and governance in the manufacturing and project management sectors. I have a continuing interest on how such development will impact the nuclear sector.
I would propose that AI’s integration into nuclear project management is not, by any means, intrinsically detrimental; rather, a nuanced and careful assessment of the integration benefits, associated risks, and effective mitigation strategies is truly warranted to ensure balanced outcomes. Artificial Intelligence has the potential to foster significant improvements across various project management dimensions, enhancing efficiency, accuracy, and decision-making capabilities; however, it is crucial to recognise that the unintended consequences of its adoption may adversely affect other essential dimensions, leading to challenges that need to be addressed thoughtfully.
This article identifies insights from the project management, artificial intelligence, nuclear, and the relevant body of knowledge to be found in academic literature. Several key constructs relative to artificial intelligence, project management quality, and the nuclear context are defined, measurement strategies are clarified, and relevant theoretical frameworks are identified. This examination foregrounds the dual perspective of integration as a potential generator of benefits and a source of risk.
In the nuclear sector safety is a paramount consideration which includes shaping project governance and driving heightened scrutiny of innovation and change. The urgency of the nuclear energy transition further underscores the relevance of artificial intelligence to the sector. Therefore, notwithstanding the possible decline in certain project management dimensions, it remains essential to understand integration through the double lens of opportunity and threat.
The nuclear context, characterised by stringent safety expectations, extensive regulatory oversight, prolonged project durations, and severe consequence potentials, shapes expectations regarding artificial intelligence integration. Artificial intelligence capabilities also intersect with sector characteristics and influence the operationalisation of project management quality. Through these, the analysis frames the investigation and clarifies the contours of the artificial intelligence-influenced governance landscape.
Conceptual Framework: AI in Project Management and Nuclear Sector Characteristics
Nuclear projects must comply with stringent, multi-tiered safety and regulatory frameworks. Given the high consequences of failure, requirements for design, licensing, and operations are critical. The project management (PM) quality framework applies to safety and regulatory decisions, covering optimal activities to fulfil requirements throughout the entire lifecycle, especially design and construction phases. The characteristics of the nuclear sector and the significance of the project PM are therefore foremost. Project delivery demands multidisciplinary expertise; high spend limits the number of competitors; and compliance with safety, quality, project delivery, or construction metrics defines success (Khanh Dam et al., 2018).
Construction disciplines which include civil, mechanical, electrical, and instrumentation engineering, dominate implementation. Regulatory controls fulfil safety mandates, permit process licensing, and govern activity acceptance (Ranjbar et al., 2024). Project engagement occurs during the front-end engineering design phase for complex energy projects (e.g., Fission, Fusion, and large-scale renewables). Certain high-dispersion algorithms based on PM, management of change, and risk management are increasingly initiated and monitored. I would propose therefore that all indications suggest that AI is not intrinsic to decline and thus, an evidence-based and objective assessment of the AI impact on the nuclear sector Project Management quality is warranted.
Potential Benefits of AI for Nuclear Project Management
There is an increasing impetus to introduce AI into project management within the sector. This is driven by the potential for AI to enhance forecasting, prescriptive analytics, and automated reasoning, thereby increasing prediction accuracy and reducing uncertainty across various dimensions. Precise schedule estimates, for example, are essential for the efficacious coordination of nuclear project governance, work-licensing interactions, and supply chain management. AI can leverage existing data on project history (eg. Learning from experience, analysis of issues registers etc), industry benchmarks, and regulatory expectations, combined with using domain constraints, to improve the time-to-schedule estimation (Khanh Dam et al., 2018).
AI techniques, such as analysis of past projects, project/task clustering, and risk cause prioritisation, can improve project planning and when integrated into a decision support system they can facilitate more rapid and effective decision making for plant management during operational projects (Obinnaya Chikezie Victor, 2023). Of particular interest to operations in my own employment, similar approaches can also assist in refining decommissioning work-programme development and in determining whether proposed extensions for operational projects would have an adverse effect on the timely implementation of decommissioning. This is a significant focus of my own research in Quality, AI and Project Management in the Nuclear and Manufacturing sectors.
Potential Risks and Threats to Quality
There are significant risks with the introduction of AI however. AI can adversely affect the quality of project management in nuclear projects. Various risks may undermine enhancements to scheduling, governance, risk management, and safety; these include miscalibration, data bias, overreliance, opacity, cyber threats, integration challenges, and regulatory noncompliance (Khanh Dam et al., 2018). Should these threats propagate throughout the project management process: malfunctions may not be detected until several phases later, when the consequences are compounded. Awareness of such potential failure modes can motivate additional assessments to avoid responsiveness limitations.
Moreover, poor discipline or high stakes can inhibit effective human–AI collaboration. Clear processes are necessary for using AI tools to make decisions on critical issues, thereby supporting regulatory compliance and safety (C. Horowitz et al., 2019). The when, how, and why AI models generate recommendations must be intelligible to serve as reliable evidence for risk-based decision making. I suggest that this clarity is essential when seeking to employ AI models in nuclear projects.
Governance, Compliance, and Safety Considerations
As the Head of Quality for a major Nuclear Contractor I would concur with the work of Ranjbar et al (2024) an Papagiannidis et al (2023) in stating that governance, compliance, and safety considerations are paramount when introducing AI to nuclear projects. Integrating AI into project management raises critical issues related to governance, compliance, and safety in the nuclear sector. To address these concerns, predefined frameworks, processes, and structures for standardisation and oversight of AI-supported project management must be established (Ranjbar et al., 2024). Such measures aim to ensure appropriate collection and processing of pertinent data, maintain traceability and auditability of data handling, define stakeholder roles and responsibilities, and adhere to safety regulations and quality standards aligned with expectations from nuclear regulatory authorities (Papagiannidis et al., 2023).
Nuclear safety represents a paramount priority throughout the project management life cycle. Consequently, all AI-enabled tools must comply with national regulations and internationally endorsed principles governing the safety of AI-based software and systems. Effective governance mechanisms help shape data management records and processes, safeguard information security, and tackle cyber threats associated with AI implementation. Consequently, a comprehensive safety case demonstrating compliance with project governance and AI system requirements is indispensable, along with alignment with the licensing and regulatory processes enforced by the relevant national authority or international body overseeing power reactor construction and operation.
Mitigation Strategies: Ensuring Quality in AI-Enhanced Project Management
Artificial intelligence does not inherently compromise quality in nuclear project management. The allocation of resources to such project management tools should be carefully considered to uphold compliance to standards and regulations. Project management quality is as critical in the nuclear sector as safety, and various safeguards can mitigate the risks associated with AI adoption.
Nuclear projects are marked by large scopes, strict deadlines, and exacting stakeholder requirements. Governance, scheduling, risk management, and safety add to the complexity. The size and duration of typical projects make quality assurance particularly challenging consequently, organisations are increasingly integrating artificial intelligence into project management. Such tools can enhance forecasting, resource allocation, and decision support, especially in uncertain environments (Khanh Dam et al., 2018). Yet while such tools may help improve project management quality, they might also degrade it. Concerns include mis calibrated models, inadequate data, overreliance, reduced human oversight, data biases, artificial-intelligence-generated content, cybersecurity threats, and regulatory misalignment. Project management quality is closely linked to operational excellence, organisational capability, and deterministic behaviors, all of which stand to benefit from enhanced compliance and accountability (Nunik et al., 2019). Previous work in academia and the sector has examined nuclear project quality, identified relevant performance indicators, and addressed both management and construction considerations. However, I would suggest that existing frameworks, however, do not examine robustly and transparently either the potential of artificial intelligence to enhance quality, or the associated risks in depth.
Quality-assurance strategies tailored to project management in the nuclear domain can help safeguard standards despite the introduction of artificial-intelligence capabilities. These might encompass governance and compliance measures, technical controls, and audits augmenting the nuclear oversight regime. A portfolio of evidence (eg. case studies, surveys, and quantitative studies) and actionable guidance for practitioners would further support the overall assessment.
Ethical, Legal, and Security Implications
According to the UN Office for Disarmament Affairs (2019), the rapid growth of the artificial-intelligence (AI)-enabled technology sector gives rise to not only great opportunities but also grave ethical and legal implications. Many industries are trying to adopt AI in certain areas of their state-of-the-art processes. The architecture, engineering, construction (AEC) sector is also investigating several AI methods and approaches, but ethical issues still remain. The AI and robotics research in the nuclear sector aims to improve planning efficiency, safety, productivity, etc. Several ethical issues appear when introducing AI tools, such as occupational replacement, construction site supervisors’ authority, regulations, etc. It is crucial to highlight and reflect on these ethical issues to better understand the legal implications of AI tools (Liang et al., 2023).
Nuclear facilities have already faced tremendous cyber threats; many cyberattacks are launched against information systems, transmission networks, control systems, measurement systems, etc. Reducing the possibility of cyber incidents and strengthening resilience against cyberattacks by introducing AI are under continuing assessment and investigation. Cybersecurity technology for nuclear facilities is regulated and highly standardised according to regulators’ enforced requirements. Therefore, safety and compliance with nuclear regulations during the life cycle of the AI model are thus critical (Ranjbar et al., 2024).
Future Directions and Research Gaps
AI has attracted considerable interest for project management and is being increasingly integrated into microstructure governance, scheduling, risk assessments, and safety analyses in the nuclear sector. Yet the potential and maturity of AI solutions, as well as their implications for project management quality, remains poorly understood (Singh et al., 2023). In turn, AI is not intrinsically detrimental to quality in nuclear project management. The gradual introduction of automation has never of itself contributed to a decrease in quality (Khanh Dam et al., 2018).
Conclusion
Assuming objective and factual consideration of Artificial Intelligence (AI) will arrive at the conclusion that AI is not intrinsic to the quality decline in project management within the nuclear sector, the evidence gathered aims for an evidence-based and thus objective assessment. Articulated here is a balanced synthesis that briefly summarises the scope of the investigation, affirms the hypotheses, and highlights the relevant findings that collectively constrain the impact of AI on project management quality in the nuclear sector.
Quality PM in the nuclear sector matters. The profound societal, environmental, and economic consequences resulting from project failure necessitate robust and sustainable project performance. Nuclear projects have traditionally suffered from cost and schedule overruns. Prompted by the rise of AI technologies, the central question explored is whether AI is inherently beneficial or detrimental to PM quality. To clarify the investigation to PM practices in the context of nuclear energy projects, AI is constrained to data-intensive types. Project management quality elaborates around cost, scheduling, nuclear safety, and preventive actions. AI therefore supports the common nuclear PM notion that AI is not intrinsically detrimental to project quality itself.
The integration of Artificial Intelligence (AI) into project management in the nuclear sector can significantly enhance the quality of project management by improving decision-making processes, optimizing resource allocation, and mitigating risks. AI's ability to analyze large volumes of data in real time allows for more informed decision-making, which is crucial in safety-critical environments like nuclear projects. Predictive analytics can foresee potential issues before they escalate, thus enhancing project safety and quality (Ranjbar et al., 2024).
Moreover, AI can improve communication and collaboration within multidisciplinary project teams. Streamlining workflows and aligning project goals enhance the efficiency of resource utilization, contributing to successful project outcomes (Hsu et al., 2021). Machine learning algorithms facilitate continuous improvement by identifying patterns and trends from past data, enabling project managers to refine their strategies for better deliverables (Hsu et al., 2021).
AI plays a critical role in risk management by providing real-time assessments and predicting potential equipment failures. This capability helps ensure timely maintenance and avoids costly delays, crucial in the high-stakes nuclear sector (Ranjbar et al., 2024). Additionally, the incorporation of AI can lead to the development of advanced control systems that enhance the operational efficiency of nuclear facilities (Massaro, 2022). These systems utilize AI to process data from different sensors, allowing for real-time adjustments and improved safety metrics.
The use of AI in project management aligns closely with the principles of Industry 5.0, which emphasises the collaboration between humans and machines to foster an intelligent workplace. In this context, AI can assist project managers in maintaining safety oversight while also handling routine tasks, thus allowing human operators to focus on more strategic decision-making (Dam et al., 2018).
However, the implementation of AI also presents ethical and regulatory challenges that must be addressed to ensure safety and compliance with industry standards. The nuclear sector, being heavily regulated, requires strict adherence to safety and ethical standards. As AI technologies evolve, so does the need for regulations that ensure their safe deployment. The integration of AI into project management processes must consider these factors to avoid potential pitfalls and ensure that the quality of project outcomes is not compromised.
Furthermore, the advent of autonomous systems within the nuclear sector presents both opportunities and challenges. Autonomous systems equipped with AI can perform inspections and maintenance tasks in hazardous environments, reducing human exposure to risk (Anderson & Dennis, 2023). Nevertheless, the development of safety cases for these systems is a critical aspect that requires careful planning and execution to ensure that AI technologies are safely integrated into existing protocols.
In conclusion, the adoption of AI in nuclear project management promises enhanced decision-making, communication, risk management, and overall project quality. However, careful navigation of ethical and regulatory issues is essential to fully realise these benefits while minimizing associated risks (Ranjbar et al., 2024; Hsu et al., 2021). Future research should focus on developing frameworks that facilitate the effective integration of AI into project management practices in the nuclear industry, ensuring that safety remains the foremost priority.
References
Anderson, C. R., & Dennis, L. A. (2023). Autonomous Systems' Safety Cases for use in UK Nuclear Environments
Dam, H. K., Tran, T., Grundy, J., Ghose, A., & Kamei, Y. (2018). Towards effective AI-powered agile project management
Grote, M. & Bogner, J., 2023. A Case Study on AI Engineering Practices: Developing an Autonomous Stock Trading System
Horowitz, C., Scharre, P., & Velez-Green, A., 2019. A Stable Nuclear Future? The Impact of Autonomous Systems and Artificial Intelligence
Hsu, M. W., Dacre, N., & Senyo, P. K. (2021). Applied Algorithmic Machine Learning for Intelligent Project Prediction: Towards an AI Framework of Project Success
Khanh Dam, H., Tran, T., Grundy, J., Ghose, A., & Kamei, Y., 2018. Towards effective AI-powered agile project management.
Liang, C. J., Le, T. H., Ham, Y., R. K. Mantha, B., H. Cheng, M., & J. Lin, J., 2023. Ethics of Artificial Intelligence and Robotics in the Architecture, Engineering, and Construction Industry.
Massaro, A. (2022). Advanced Control Systems in Industry 5.0 Enabling Process Mining.
Nunik, M., Berawi Mohammed, A., , G., & I Gede, S., 2019. Dominant factors influencing project quality in the radioactive minerals processing pilot plant construction.
Obinnaya Chikezie Victor, N., 2023. Impact of Artificial Intelligence on Electrical and Electronics Engineering Productivity in the Construction Industry.
Papagiannidis, E., Merete Enholm, I., Dremel, C., Mikalef, P., & Krogstie, J., 2023. Toward AI Governance: Identifying Best Practices and Potential Barriers and Outcomes.
Ranjbar, A., Wermundsen Mork, E., Ravn, J., Brøgger, H., Myrseth, P., Østrem, H., & Hallock, H. (2024). Managing Risk and Quality of AI in Healthcare: Are Hospitals Ready for Implementation?
Singh, A., Dwivedi, A., Agrawal, D., & Singh, D., 2023. Identifying issues in adoption of AI practices in construction supply chains: towards managing sustainability.
Zapke, M., 2019. Artificial intelligence in supply chains.