Quantitative Evaluation of AI Productivity and Quality in Design Processes: A Case Study on Engine Piston Diameter Calculation and 3D Modeling

Artificial Intelligence (AI) is currently regarded as the foremost technology for automating routine and repetitive tasks, leading to increased productivity. However, the quality of creative and design work with AI remains questionable. This paper presents a quantitative analysis of AI productivity through dynamic simulation and assesses the quality of AI results in the diameter calculation and construction of a 3D model of an engine piston as a case study. To evaluate productivity, the dynamic model segregates design tasks based on AI working hours. The quality of the formulation for calculating the engine piston diameter, derived from engine requirements, is compared with a standard formulation from a literature review. Additionally, the 3D model generated by AI is compared with a model created by human intelligence in Computer-Aided Design (CAD) software, reflecting the characteristics and properties of real engine pistons. While research on AI productivity is abundant, few studies address the quality and usefulness of AI-generated results. This study aims to evaluate these three aspects. As anticipated, the AI in a simulation model demonstrates a numerical increase in productivity as an enhancing variable. However, results for a design process involving mathematical formulation and 3D model construction lack utility without additional work. Our findings lead us to conclude that AI in the design process can enhance productivity when used to suggest and predict design instructions, thereby saving time. Nevertheless, the AI's ability to create mathematical and 3D models is limited to simplified conditions, and further knowledge must be imparted to the AI to enable it to produce readily usable designed components.