Innovation through pertinent patents research based on physical phenomena involved

One can find innovative solutions to complex industrial problems by looking for knowledge in patents. Traditional search using keywords in databases of patents has been widely used. Currently, different computational methods that limit human intervention have been developed. We aim to define a method to improve the search for relevant patents in order to solve industrial problems and specifically to deduce evolution opportunities. The non-automatic, semi-automatic, and automatic search methods use keywords. For a detailed keyword search, we propose as a basis the functional decomposition and the analysis of the physical phenomena involved in the achievement of the function to fulfill. The search for solutions to design a bi-phasic separator in deep offshore shows the method presented in this paper