Catching Up with Expert Systems

century, a new era that will be far more scientific, technologic, and sophisticated than anyone would have imagined just a quarter of a century ago. However, the continued success in all areas of pharmaceutical science will depend entirely on how fast pharmaceutical scientists will adapt to rapidly changing technology. Almost 10 years ago, a survey by Shangraw and Demarest revealed a very interesting fact about solid-dosage formulation design and development: Tradition was still a very important reason for preferring to use a particular excipient (1). It is not difficult to predict that, in this century, trial and error formulation development and traditional excipient selection will be a part of history. Pharmaceutical formulators will enjoy the availability of the harmonized and fingerprinted (in terms of functionality testing) excipients, and formulations will be developed using databases (preformulation and compaction data banks, etc.) (2). The awareness of and the use of artificial intelligence–based expert systems (rule-based systems, fuzzy logic, genetic algorithm, artificial neural networks [ANNs], simulations, etc.) in the areas of preformulation, formulation and process development, regulatory affairs, new drug delivery system development, project management, and all other areas of pharmaceutical science will increase dramatically. To shorten the adaptation period of pharmaceutical scientists to rapidly changing technological advances, I recently formed two new focus groups, namely, the Expert Systems Focus Group and the Excipients Focus Group. They have been approved by the American Association of Pharmaceutical Scientists (AAPS). They will act in conjunction with the Pharmaceutical Technology Section of AAPS and are open to all members of AAPS and other pharmaceutical associations. I would like to give a brief overview of expert systems (ESs) and then address some challenges facing ES developments in terms of their verification and validation (V&V) processes, in part because of FDA’s interest in the V&V of all types of software. In a future article, each of the following issues will be discussed in depth. ESs, also known as knowledge-based systems, basically are computer programs that either recommend or make decisions based on knowledge gathered from experts in the field. Functional areas of ESs include, but are not limited to, control, design, diagnosis, instruction, interpretation, monitoring, plan-