Abstract
<title>Abstract</title> <p>The transition from theoretical thermodynamics to practical fluid and thermal engineering demands a precise understanding of non ideal boundary conditions. Undergraduate and postgraduate mechanical engineering curricula heavily rely on classical laboratory experiments to demonstrate fundamental mechanisms like steady state conduction and forced convection. Students frequently struggle to reconcile the exact mathematical solutions taught in the classroom with the physical realities observed in the laboratory. The present study evaluates the educational and experimental outcomes of hybridizing classical thermal experiments with modern predictive modeling methodologies. By comparing empirical data from standard heat transfer apparatuses with theoretical textbook calculations and machine learning frameworks (including Artificial Neural Networks, Support Vector Regression, and Random Forest models), this research proposes a modernized framework for mechanical engineering laboratories. To update the analytical work flow, all computational modeling, data logging, and manuscript generation were executed entirely on a mobile tablet architecture. This approach establishes that complex thermodynamic simulations and data science pipelines no longer strictly require heavy desktop simulation hardware. Experiential learning, coupled with accessible mobile computational tools, significantly accelerates student comprehension of transient state limitations and secondary heat leakages in physical thermal systems.</p>