A Generic Simulation Model that Reflects the Flexibility of an Automated System for Pharmaceutical and Chemical Laboratory Testing

Merck & Co., Inc. has developed a laboratory automation system that is targeted towards low volume pharmaceutical and chemical testing. The nature of laboratory testing requires this system to adapt to changes in assays and testing procedures, as well as a variety of laboratory and material handling device characteristics and configurations. This adaptability is essential to keeping the laboratory from becoming obsolete. While low volume efforts are not generally viewed as opportunities for automation, this system's design, incorporating modular laboratory equipment cells and robotic material handling devices under a common controller, provides the requisite flexibility necessary for laboratory adaptation while retaining the advantages offered by automation. Inherent to this system's flexibility is the recurring problem of identifying the desired or optimal laboratory configuration for a given implementation instance when faced with several choices. While discrete-event simulation has proven to be quite beneficial for solving problems of this nature, parochial design philosophy has tended to result in models that are brittle in structure. The focus of this paper is a discussion of a single generic simulation model written in the SIMAN IV simulation language that emulates the adaptability of the flexible automation system it represents. Like the automated system, the model allows for convenient changes in nearly all aspects of the process being studied. Equipment and operations within the model may be easily included (or excluded) from the scope of a given simulation run. Included within this discussion is a description of the Merck Flexible Laboratory Automation System as well as a case study applying the flexible simulation model to an actual laboratory instance.