With the rise of lab automation technologies, scientists are writing programs instead of pen-and-paper protocols. These programs mix computation with fluidic manipulation, leading to problems with resource management and error handling. Scientists have rich understanding of the biological or chemical systems involved, but the current programming model can not use this information. We believe techniques from the programming language (PL) community are well-suited to address the myriad challenges in this field. This talk will present how some of the problems in this field line up with (and challenge) well-understood PL techniques.