(564f) Practical Applications of Generative Artificial Intelligence in Process Simulation

Artificial Intelligence (AI) has been consolidating as a powerful, palpable, and widespread tool in many areas. AI is revolutionizing, and potentially surpassing, some human capabilities, sparking both visionary predictions and concerns among leading figures [1-4]. Recently, AI technologies, especially deep learning, have made significant strides in processing existing data and generating new designs, setting them apart from conventional expert systems and biologically inspired algorithms. Generative AI design, which utilizes machine learning algorithms to learn from data and create new content, has been recognized as one of the top ten technology trends for 2023 [5]. Intelligent generative technologies such as, ChatGPT by OpenAI, Gemini by Google [7], Mistral by MistralAI, Claude by Anthropic and LLaMA by Meta AI have demonstrated the versatility of generative AI across various fields, establishing it as a leading, and in some cases, a priority area of research. In this context, this kind of environment can help engineers navigate through many design processes based on architectural plans with multiple constraints such as compliance and cost-effectiveness. Thus, the end-to-end design process of generative AI mirrors that of structural engineering, equipped with robust learning and generative capabilities to address the complexities of intelligent design [8]. This journey, in our work, begins with a foundational prototyping phase, where an initial simulation is created based on the principles of GPT's learning derived from the AVEVA Process Simulation (APS) Scripting, a specific tool that allows direct communication between APS and programming language as Python or C#. This initial step is a test related to examining GPT's cognitive abilities in understanding the complex codes and processes inherent in the APS-Python communication domain. This early stage is of paramount importance, revealing a range of challenges and opportunities while also validating the technical and practical feasibility of the project [2]. The initial results have demonstrated that this integration goes beyond simply demonstrating AI's effectiveness in streamlining process integration, heralding a new era of innovation in process engineering. The integration of AI with advanced simulation offers numerous benefits, including reducing operational costs and improving industrial efficiency while adeptly handling the complexities and variabilities of production environments. The combination of AI methodologies based on GPT will redefine the landscape of process integration and optimization within industrial settings, giving process engineers the possibilities to develop deeper insights and more specific applications where human creativity is needed. In conclusion, the development of tools that facilitate the initial steps of process simulation is essential in the modern engineering landscape. By automating routine tasks and streamlining the integration of AI and simulation platforms, engineers are liberated from mechanical procedural work. This freedom allows them to focus their efforts on more creative and innovative aspects of process design and optimization. Such tools enhance efficiency and productivity resulting in advancements in process engineering, where human and AI-driven insights converge to shape the future of industrial processes.