The considerable improvements in computational power and the increased sophistication of programming techniques now make it possible to automate many aspects of electron microscopy [1]. Automation offers several important advantages over manual operation. Automation of microscope operation permits novice microscopists to obtain data relatively easily thus increasing instrument utilization. The experienced microscopist should be able to acquire data more rapidly and with greater ease compared to manual operation, resulting in increased efficiency. Alignment of energyfilters and proposed Cs correctors are difficult and time consuming even for experienced microscopist and stable alignment algorithms are critical for practical use. Automation is essential for extensive statistical sampling of microstructures and can permit experiments to be performed that were previously impractical. For example, spectrum imaging involves acquisition of thousands of spectra from a predefined area. The acquisition and processing of such large volumes of data cannot be accomplished without the aid of automation. The future prospects for automation are exciting, although it is important to carefully consider the relationship between the computer controlled microscope and the operator. Here I give a personal perspective on approaches to automation which increase the power and efficiency of modern transmission electron microscopes.

The degree of automation is an important consideration that is related to the complexity of the analysis strategy. Some analysis situations call for repetitive sampling of a well-defined quantity such as layer thickness or particle size. These simple repetitive operations are ideal candidates for total automation. For many applications of automation it is necessary not only to acquire the data but also to process the data and extract the quantity of interest.