Shaft misalignment and rotor unbalance are the two major concerns in modern industry. A novel approach to condition monitoring of machinery known as Coast Down Time (CDT) is being studied to use as a diagnostic parameter for detecting and analyzing the effect of misalignment and unbalance under experimental and simulated industrial environment. Further, the use of Artificial Neural Networks in CDTs prediction for different mechanical fault conditions at various levels, fault detection and classification are also investigated. An analytical expression is developed for CDT estimation and compared with the experimental CDT results. Industrial case studies were carried out to validate the experimental data; the results are found to be well in agreement with theoretical and experimental CDT results.