It is often considered that there are two types of scientist: theoretical and experimental. Theoretical scientists invent/discover new theories to explain the natural world. Experimental scientists invent experiments to test these theories. For example, the theoretical physicist Albert Einstein invented the theory of General Relativity, and astronomers tested General Relativity by inventing the idea of observing the position of stars near the Sun during a solar eclipse. Artificial Intelligence (AI) is increasingly being used in scientific research. Almost all of this effort is in the formation of new hypotheses to explain data. This corresponds to what theoretical scientists do. Far less effort has been put into automating what experimental scientists do. This is what the proposed research focus on. We propose to develop a 'Robot Experimentalist'. This will be an AI system that when given a hypothesis to be tested, and a description of a set of laboratory equipment, will be able to plan an experiment to test the hypothesis. To make the problem tractable we will focus on experiments using the yeast S. cerevisiae. This is the organism used to make bread, beer, and wine; but its main role in biology is as a model for human cells. Surprisingly, most of what is true for S. cerevisiae is also true for H. sapiens. We will test the Robot Experimentalist using laboratory robots in the UK and Japan.