Abstract
The purpose of this article is to report on a systematic review performed on the literature describing the research on the use of big data technologies in pharmaceutical manufacturing. Big data technologies refer to a variety of technologies that are associated with big data such as data analytics, machine learning, and data mining. The systematic review uses a set of search terms that describe the topic under study and performs searches on major databases. The returns from these searches are subjected to content analysis to determine their primary topic and those that are focused on pharmaceutical manufacturing are then analyzed to reveal research emphasis, tools used, and major findings. The literature is then organized to reveal the structure of the body of literature. A total of 64 papers met the inclusion requirements for the study and they represent four themes: Manufacturing Specific Products, General Manufacturing Process Control, Safety/QC, and Manufacturing Management. A review such as this provides guidance to researchers to expand their current research.
TopBackground
This section provides more specific definitions for the concepts under study in this paper. The first is the term “pharmaceutical manufacturing.” For the purposes of the current study, this term applies to the processes used to actually create the drugs and compounds in whatever form they might take (tablets, liquids, etc.) as well as the ancillary operations that are necessary to the success of these processes. These might include, but are not limited to, quality control, safety monitoring, material procurement, etc. In short, whatever aspects of pharmaceutical manufacturing that is returned using the search terms detailed in the next section will be considered. It is not the purpose of this study to decide which activities are considered to be part of the manufacturing process, but rather to report on what the researchers and the practitioners in the field consider them to be.
Key Terms in this Chapter
Datamining: A category of tools that are able to sift through large and sometimes disparate datasets to find trends or categorize the data into subsets that are more descriptive of the concepts buried therein.
Chemometrics: The application of mathematics, statistics, and other data science tools to extract information from data describing chemical properties or interactions.
Systematic Literature Review: A systematic approach to describing the literature surrounding a well-defined discipline or sub-discipline.
Deep Learning: A subtype of machine learning that refers to tools such as Artificial Neural Networks (ANN). These tools are trained to make decisions such as classifications or speech recognition. These are considered to be related to big data because the datasets required to train and verify the algorithms commonly fall into the general definition of big data.
Data Analytics: A broad term covering all of the statistical and mathematical tools and techniques commonly used to analyze and draw meaningful information from big data repositories.
Machine Learning: A broad category of algorithms that can be trained to create the ability to make decisions. There are many different types of training methods and architectures, a full discussion of which is out of the scope of this paper.
Knowledge Management: The application of technology and organizational processes to promote the capture, classification, storage, and application of organizational knowledge to create value for the organization.