ABSTRACT
Autonomous intelligent decision-making might just be the core of the autonomous loop and certainly represents the difference between automation and autonomy. While automation means no human interference is required for the experiment execution, autonomy implies a sense of artificial intelligence that is adapting to what is being learned during the data acquisition and as a result, is changing strategies throughout the experiment. The idea behind Gaussian Processes is to think of function values of a model to be jointly normally distributed. In most situations the authors come across when attempting an autonomous experiment, they have a clear understanding of what parameters can be changed, which directly translates into the parameter space they want to explore. That space can be made up of motor positions, temperatures, atmospheric pressure, fluid ratios, and so on.
