Artificial Intelligence in Mechanical Engineering: Foundations, Opportunities, Challenges, and Responsibilities
Authors: A. Cunha, C. Nataraj, L. M. Lorenzo, H. Panetto, A. Scotti, J. V. C. Vargas
Organizations: National Laboratory of Scientific Computing, Av. Getúlio Vargas, 333, Petrópolis, 25651-075, RJ, Brazil · Rio de Janeiro State University, Rua São Francisco Xavier, 524, Rio de Janeiro, 20550-900, RJ, Brazil · Moritz Endowed Professor, Villanova Center for Analytics of Dynamic Systems (VCADS), Villanova University, Villanova, PA 19085, USA · Dassault Systèmes, Av. Jornalista Roberto Marinho, 85, São Paulo, 04576-010, SP, Brazil · University of Lorraine, CNRS, CRAN, Nancy, F-54000, France · Federal University of Uberlândia, Rodovia BR-050, km 78 - Bloco 1DCG, Uberlândia, 38410-337, MG, Brazil · Federal University of Paraná, CP 19011, Curitiba, 81531–980, PR, Brazil
Artificial intelligence (AI) is reshaping mechanical engineering (ME), from modeling, simulation, and optimization to diagnostics, design, and technical communication. Drawing on a round-table at the 28th ABCM International Congress on Mechanical Engineering (COBEM 2025), this article presents a structured perspective for the Brazilian and international ME communities. Rather than an exhaustive survey, it examines how AI can enhance engineering practice without compromising scientific rigor. We distinguish statistical learning from human cognition and discuss the capabilities, limitations, and emerging agentic forms of AI. We then consider industrial adoption, workforce changes, and AI-assisted scientific discovery, alongside the energy and infrastructure demands of AI computation and their implications for sustainability and thermal management. The discussion emphasizes the integration of physical knowledge, structured reasoning, and domain expertise into AI systems, particularly in safety-relevant applications. Topics include physics-informed and hybrid learning, physically constrained modeling, validation, uncertainty quantification, and certification. We also examine educational, epistemological, and ethical challenges, including curriculum reform, human-agent collaboration, professional accountability, and the preservation of engineering knowledge. We conclude with principles for responsible AI adoption in ME: physical consistency, rigorous validation, explicit treatment of uncertainty, energy sustainability, human oversight, and ethical responsibility. AI should complement, not replace, engineering judgment.
Figures & tables
Figure 1: System-level view of artificial intelligence in ME. Engineering applications are enabled by AI methodologies, including machine learning, physics-informed learning, generative models, and emerging agentic systems that coordinate simulations, optimization, and data-driven workflows. These capabilities remain grounded in physical laws, domain knowledge, experimental evidence, energy infrastructure, and governance frameworks. Validation, uncertainty quantification, certification, and traceability act as cross-cutting requirements across all layers, ensuring that AI systems can be deployed responsibly in engineering practice. The diagram emphasizes that AI functions as an embedded component of engineering systems rather than as an autonomous substitute for engineering judgment.
Figure 2: Conceptual architecture of a human-agent engineering workflow. Engineering objectives are pursued through collaboration between a human engineer and an agentic AI system, which interact with an ecosystem of computational tools including CAD models, physics solvers, optimization engines, engineering databases, sensors, and digital twins. The agentic AI system assists with planning, reasoning, information retrieval, and workflow orchestration, while the human engineer provides domain expertise, physical reasoning, engineering judgment, validation, and accountability. Intermediate results generated throughout the workflow undergo verification, validation, and uncertainty quantification before supporting engineering decisions and deployment. Physical constraints, governance requirements, and ethical considerations act as cross-cutting influences across all stages. The diagram emphasizes that agentic AI can coordinate complex engineering workflows, but trustworthy deployment ultimately depends on human oversight, physical consistency, uncertainty awareness, and professional responsibility.
Figure 3: Schematic representation of the energy flow and infrastructure requirements of AI-driven data centers. Electrical power from primary sources is delivered through distribution and conditioning systems to high-density computing hardware (GPUs, CPUs, and networking equipment), which generate substantial heat loads. Thermal management subsystems—including liquid- and air-cooling—require additional electrical power, thereby increasing total facility power demand and shaping the scalability and sustainability of AI infrastructure.
Figure 4: Conceptual framework for AI-enabled ME education. Future engineers must integrate four complementary domains of competence: (i) strong engineering fundamentals grounded in physical laws, modeling, and analysis; (ii) artificial intelligence and data-driven methods; (iii) ethical awareness and professional responsibility; and (iv) human-agent teaming and agentic workflow supervision. The latter includes competencies such as physics-aware interaction with AI systems, supervision of AI-assisted workflows, inspection and validation of AI-generated analyses, and critical assessment of intermediate and final computational results. The integration of these domains enables the responsible, effective, and trustworthy use of AI in engineering practice.
Responsible AI studies how to design, evaluate, and govern AI systems to promote fairness, safety, transparency, accountability, privacy, and sustainability.
McKetta Department of Chemical Engineering and Oden Institute for Chemical Engineering and Sciences, The University of Texas at Austin, Austin, TX 78731, USA · Department of Chemical and Biological Engineering, Northwestern University, Evanston, IL 60208, USA · Department of Chemical and Biomolecular Engineering, University of Delaware, Newark, DE 19716, USA +13