You are here

Automated Experimentation as a Means to Promote Computational Thinking in the Sciences

e-Science Institute Public Lecture
Dave Robertson

We already have glimpsed what computation can do for the traditional sciences when used as a tool in support of conventional experiments. The tools that have been built, or soon will be built, allow experiments at a scale and intensity that were difficult to imagine even a few years ago. This advance, however, has created a problem that grows as the use of powerful computational tools becomes more prevalent : the nature of the experiments themselves change as they come to rely more on automation and scientists must (at some level) understand computation in order to trust their own experiments. The question is, what to do about this? An answer will require us (that is, traditional scientists and informaticians) to identify the essence of how one should think computationally about specific scientific domains and, through this, develop the forms of education and automation that allow science to flourish in these new arenas. I shall discuss this issue using, as concrete examples, experiments using protocols and distributed workflows.

Date and time: 
Tuesday, 23 February, 2010 - 16:00
60 minutes