Engineering, Procurement, and Construction (EPC) projects rely heavily on detailed engineering data and accurate 3D models. Artificial Intelligence (AI) opportunities offer a transformative vision for this process, promising automation, optimization, and improved collaboration. Also, AI is being developed to build 3D models of process plants with multiple design scenarios augmenting human knowledge. However, integrating AI in FEED Engineering workflows comes with several challenges: Integration and collaboration of data: EPC projects involve numerous stakeholders with diverse data formats and software tools. Ensuring seamless data exchange and interoperability between AI-powered modelling platforms and existing software used by engineers, designers, and fabricators is crucial for such integration.Engineering Data Quality: AI models need to be trained on data that reflects specific engineering design requirements, codes and design practices relevant to each EPC project or process package. As part of this challenge is validation and verification of data. The complex nature of EPC projects necessitates robust validation processes to ensure the accuracy and efficiency of AI-generated models.The AI "black box": Capturing the design intent of specific design and modelling options or decisions is crucial for EPC projects. The "black box" nature of some AI algorithms can make it difficult to understand the rationale behind generated designs or gain support from engineering organizations for company wide deployment.Cultural Resistance to AI: Cultural resistance to AI solutions stems from hesitation, scepticism, or outright opposition that some individuals or groups within an organization may have towards adopting and implementing Artificial Intelligence (AI) technologies.
This paper seeks to raise awareness of the challenges recognized by current literature in the industry and discuss opportunities for implementing AI solutions when developing 3D models for FEED projects. This paper will also propose best practices for harnessing the strengths of AI to optimize its benefits. The following key areas are discussed as AI opportunities on EPC projects during 3D modelling in FEED: Data Quality, Integration and Standardization: Ensure data used to train AI models is accurate, reliable and follows standardised formats throughout the EPC project lifecycle for seamless data exchange. Inconsistent data can lead to unreliable AI outcomes.Verification and Validation: Develop robust verification and validation processes to ensure the accuracy, quality, safety, and constructability of AI-generated 3D models.Integration of human knowledge and AI tools: While AI automates tasks, human expertise remains crucial. Integrate human oversight throughout the process for design intent capture, validation of AI outputs, and final decision-making.Develop Human skills: The effective use of AI in EPC projects requires a workforce with a blend of engineering expertise and AI skills. Developing an AI culture within the organization and investing in training programs that embraces human-AI collaboration is critical.
By implementing the AI opportunities outlined in this paper, EPC projects can harness the advantages of AI to enhance efficiency, reduce costs, and improve project outcomes. Engineering firms should also focus on empowering and developing their employees with AI skills to foster collaboration between human expertise and AI tools, while addressing the cultural concerns surrounding job security. AI has the potential to serve as a powerful tool for automation, optimization, and collaboration of data during the full project lifecycle, revolutionizing the design and construction of complex engineering projects.
