The first step was to define the lab's requirements and select a solution. Those discussions spanned several months, and resulted
in the selection of a distributed, hardware-independent process automation system. A comparison of manual and automated is
in Table 2.
Table 2. Comparing manual and automated controls
PLANNING A CHANGE
Our automation team worked with the Lab Director to specify the requirements for his Process Science labs. The life science
industries are late when compared to the electronics and other manufacturing industries in adopting automated process control
and data management systems. Biopharmaceutical firms are aware that huge profit opportunities exist for those who can get
higher-quality, higher-yield products to market faster. Automation of process development at these facilities—from a small
academic lab to a biotechnology giant— must be much more flexible and affordable than traditional manufacturing automation
Discussions with the Process Science group lasted several weeks. During this time, the group defined their requirements and
evaluated different solutions. They listed the instruments they would be using, specified some of their control methods and
analysis, and talked with their information technology (IT) department about network and backup plans.
The Process Science group selected an instrument-independent, distributed process control and data management software system,
specifically designed to increase productivity and accelerate product development in cell culture and fermentation. By tying
together existing instrumentation with comprehensive software, they brought the efficiency and accuracy of automation to their
labs, and allowed the Lab Director to save in his capital equipment budget by making full use of the legacy and acquired instruments.
The automation system this company selected is very cost-effective. Most systems with comparable features and benefits would
be more expensive, and would require specific instrumentation to be used: either a particular brand of biocontroller or expensive
PLCs. Such systems require months of expensive customization to get a facility up and running. Instead, this software solution
has a full range of capabilities yet it is hardware-independent, cooperating with biocontrollers, PLCs, and other instruments
as desired. It is quick to install and configure, and it is specifically designed for fermentation and cell culture.
There is cheaper software available from individual instrument vendors. While this software typically works well with the
specific instrument for which it was designed, it often lacks the robust features, networked architecture, open integration,
and flexibility needed to provide automated process control and data management for the entire lab.
The software was installed in their bacterial fermentation labs on standard Windows desktop PCs, plus one Windows server-class
computer, all on a standard local area network. The different brands of biocontrollers were connected to the networked computers
via Ethernet or serial ports. The other instruments were connected in various ways: Masterflex and Watson-Marlow pumps via
USB data acquisition cards; Sartorius mass flow controllers integrated with the Braun Biostat B reactor controllers; Mettler-Toledo
and Ohaus balances to serial ports; and dissolved oxygen and optical density probes through FieldPoint modules.
The total installation and configuration took one day, followed by two days of on-site training. Since then, further configuration
and extensions—such as new instruments, control methods and displays — have been added over the years without requiring system
After using the software for some time, the Lab Director reported that the automation system "...provides flexibility, customization,
data analysis, and data review that are very accessible and very well done." He described the earlier manual methods as "inefficient
and error prone." The system was soon extended into the cell culture lab. This added a dozen Applikon ADI 1030 biocontrollers
to the existing array of instrumentation under its control. Process scientists at the company today continue to extend the
automation system as needed, modifying and adding new recipes and experiments; measurements and calculations; control methods;
and reports and displays.