Embedded e-diagnostic for distributed industrial machinery

Industrial process machine failure often causes severe financial implications. This is compounded by the lack of availability of experts and the complications of getting them to site. One solution is to give the expert access to the machine remotely with the addition of an Artificial Intelligence (AI) based diagnostics software to assist with the decision making process. Our research is based on such a system, which combines modern communications with intelligent diagnostics software. Accessibility to process machines can now be global with the promise of predictability to the diagnosis. It is felt the importance of this research work cannot be overstated with the constantly moving worldwide manufacturing base and the real situation of the machine designers being based in a different country to their customer. The most vulnerable areas of a machine are its parts that consist of electro-mechanical actuation. The author utilises conventional Newtonian physics and differential calculus to model these and an AI technique of fault prediction and detection.