Artificial neural networks (ANNs) are good at approximating complex and non-linear data. In addition, they have excellent predictive capabilities and can be configured to be self-adaptive. As a result of these characteristics, the potential applications of ANNs are many and in diverse fields. These range from predicting the output of a manufacturing process through differentiating between handwritten letters to predicting the winner of a horse race. In this paper, we focus on applications of artificial neural networks to thermal systems including chemical vapor deposition, thermal management and heat exchangers.