Integrating laboratory robots with analytical instruments--must it really be so difficult?

Abstract

Creating a reliable system from discrete laboratory instruments is often a task fraught with difficulties. While many modern analytical instruments are marvels of detection and data handling, attempts to create automated analytical systems incorporating such instruments are often frustrated by their human-oriented control structures and their egocentricity. The laboratory robot, while fully susceptible to these problems, extends such compatibility issues to the physical dimensions involving sample interchange, manipulation, and event timing. The workcell concept was conceived to describe the procedure and equipment necessary to carry out a single task during sample preparation. This notion can be extended to organize all operations in an automated system. Each workcell, no matter how complex its local repertoire of functions, must be minimally capable of accepting information (commands, data), returning information on demand (status, results), and being started, stopped, and reset by a higher level device. Even the system controller should have a mode where it can be directed by instructions from a higher level.