Below is a reference to a great summary article concerning laboratory automation within the pharmacogenomic field.

Kimimichi Obata‌, Hideji Tajima‌, Masafumi Yohda‌, Tadashi MatsunagaRecent developments in laboratory automation using magnetic particles for genome analysis. Pharmacogenomics
September 2002, Vol. 3, No. 5, Pages 697-708
doi:10.1517/14622416.3.5.697

Abstract:
The majority of research for genome analysis has shifted from nucleic acid sequencing to the biological functional analysis of each gene. Based on past success, it may not be long before genome diagnostics becomes a widespread tool in human, veterinary and botany research fields. Genome analysis involves the processes of nucleic acid purification, amplification, labeling and signal detection (specific reaction, separation and signal counting). Except for the purification of nucleic acids, the other processes cannot be achieved without instruments, resulting in the advancement of automation processes. Since purification of nucleic acids can be done manually, automating this process has been delayed. However, because the purification of nucleic acids using magnetic particles is suitable for automation, its development has also been accelerated. The need for full automation for other processes is not as great because the majority of genome analysis is to identify the nucleic acid sequence and analyze genome expression. However, once useful diagnostic tools are generated, the desire for full automation will significantly increase. In order to develop realistic and practical automation, various technologies developed for each process in genome analysis have to be evaluated and only a few technologies, useful for automation, selected. The other key factor in automation is the development of methods to manage reagents and reaction mixtures precisely without any risks specifically related to genome handling, such as cross-contamination. Methods using magnetic particles, which have been used for the automation of nucleic acid purification and immunoassay, appear to be the most promising way to automate processes used in biological research.