Expert System In The Development Of Pharmaceutical Formulations

Sponsored Links

Tejas Thakkar

Tejas Thakkar

It is well known that in an Information Technology (IT) driven society, knowledge is the most significant asset of any organization.  The role of IT in health care system is well established.1.

In last few decades, tremendous advancement and development has been made by utilizing IT in the field of pharmaceutical formulation development. Preparation of various dosage forms such as oral (tablet, capsule), parental (intravenous injection) or semisolid (skin based products) etc. has begin with product specification and end up with the final recipe of formulation. In development of the new formulation may require list of ingredients with different proposition. Performance and specification of formulation varies with changing the proposition of excipients and processing variables. At the time of developing new formulation, formulator may account the physical, chemical and biopharmaceutical properties of active drug as well as possible chemical interaction between added excipients which improve the performance of final formulation. Excipients and its proportion are fixed by performing laboratory level experiment and routine database of past experiments. The performance criteria of developed formulations are need to refined in the light of experience of experts. Process of formulation development is beginning with chemical synthesis of new substance to the entrance of product into market which may take 12 years and cost $400 million2. So it is need of era to reduce cost of R&D and save time during development processes.

Pharmaceutical formulation development is an information and knowledge intensive process.  By use of new Process Analytical Technology (PAT), scientists has enables to gets a better understanding of underlying physical and chemical phenomenon. The knowledge created from the learning processes can be in different forms like reports on paper, scientific hypothesis, journals, official guideline, mathematical models, experience gained by scientist, experts and experience person. Therefore more information and knowledge become available; it is needed the more powerful and intelligent software system to mange and accesses all available scientific data for efficient decision making process. To support the activity and decision making task in the formulation development process, it may  required a systematic integrated frame work which based on formal and explicit modeling of related information. Various form of knowledge including heuristic rules, guidelines, evidence based data and mathematical models are needed to be handled in a systematic manner so that knowledge can be easily created, use independently or in an integrated fashion. By using such frame work, expert decision support systems can be developed to provide decision support prospectively. Several expert systems have been developed by numbers of pharmaceutical companies and academic institutes. This article describes concept of expert system used in development of different type of formulations. 5

Expert System 6, 7

“An expert system is a computer program that draws on the knowledge of human experts captured in a knowledge base to solve problems that normally require human expertise.”

“An expert system is a knowledge-based system emulates expert thought to solve significant problems in a particular domain of expertise.”

In its simplest form, an expert system has three major components: An interface, monitor and key board that allows two way communications between the user and the system;

1)A knowledge base where all the knowledge pertaining to domain is stored; and 

2)An inference engine/ Reasoning engine where the knowledge is extracted and manipulated to solve the problem at hand.

Inference engine strategies may be either 8,9 forward chaining, which involves the system reasoning from data and information gained by consultation from the user to form a hypothesis, or backward chaining, which involves the system starting with a hypothesis and then attempting to find data and information to prove or disprove the hypothesis. Both strategies are included in most expert systems.

Basic Components of Expert System

Figure 1. Basic Components of Expert System


Knowledge in any domain takes the form of facts and heuristics. Facts can be valid, true and justifiable by the rigorous argument while heuristics may be considered as a rule of thumb, being expert’s give best judgment in any particular circumstances. Associated with these are the terms, data and information, the former referring to facts and figures, the latter being data transferred by processing  such that the data are meaningful to the person receiving the information so we can say that knowledge is regarded as information combined with heuristics and rules . 

Acquisition of knowledge is a most difficult, time consuming and tedious aspect of expert system.  Some time it becomes difficult to manage. However, it is one of the most important elements in design of expert system. In another sense, the basic model of knowledge acquisition is the intermediate between the experts, users and knowledge bases. The knowledge engineer can acquire knowledge not only from the experts but also from all other potential sources such as written documents (like research reports, reference manual, operating procedure, policy statement etc.) as well as consultants, users and managers. In case of experts, knowledge is acquired through face to face interview. It demands a great understanding between experts and knowledge and expert engineer.

Mostly knowledge is acquired through rapid prototyping approach. In this approach, the knowledge engineer is developed a small prototype system as early as possible.  Finally, it is validated by the experts and users with several modification and addition. These methodologies are more successful in development of pharmaceutical formulation expert system.

Once the knowledge is acquired, it can be represented as a knowledge based includes the production rules, frames, semantic networks, decision table, and trees and objects.10 Probably the methodology of production rule is expressed with the relationship between the information and represented the conditional statement for the specific situation. 

Expert systems can be developed by using conventional computer languages, special purpose languages or with assistance of development shells or tool kits. Conventional language such as PASCAL and C has more applicability and more flexibility to create the control and inferencing strategy is required. Sometime the specialized languages such as LISP, PROLOG and SMART TALK have been used extensively in the development of expert system. These specialized languages are wide applicable, flexible, and very faster in implementation. As compare to other conventional languages.

Expert system shells and tool kits are the set of computer programmes written in conventional or specialized language that can create an expert system when the desirable knowledge is loaded. They offer basic facilities, including the means to prepare and store knowledge as a set of rules and two make deductions by chaining the rules together in an inferential process. Shells have different secondary characteristics such as user interfaces, operating speed, the method of knowledge presentation and associated algorithmic and arithmetic computational facilities. Therefore, computer technology in the form of expert system provides an affordable means of capturing these knowledge and expertise in a documented form that available to all.

Steps involved in pharmaceutical formulation development 11, 12

The processes of formulation development for different type of dosages forms are almost same. As figure: 2 shows different stages of formulation development such as chemical synthesis of drug, chemical and physical characterization, preformulation study, development of formulation, preclinical and clinical trials, certification and formulation marketing.

Stages involved in development of pharmaceutical formulation

Figure 2: Stages involved in development of pharmaceutical formulation

Development of formulation is highly multivariate task. There are many dependencies among drug substance, excipients and procedure. Before developing newer formulation, formulator has to consider quantity and properties of ingredients as well as possible interaction between the ingredients. A manufacturing process adopted to produce the dosage form and key operating conditions of the process may play significant role in the final quality and performance of the dosage form. Experts are frequently adept at navigating through the design space; however, their knowledge and thought processes are difficult to quantify, explain and to transmit. This kind of expertise system may helpful to senior formulator to train the new personal. In addition, retirement or personal move can lead to a loss of irreplaceable commercial knowledge that could be compensated by use of the expert system.

Benefits of Expert systems 13, 14:

(i)Knowledge protection and availability.

The existence of coherent and durable knowledge base is not affected by staff turnover. Number of experts systems is helpful for easy availability of information and the rapid access of physical and chemical data of drug and excipients which reduce the time spends for searching the literature.


All system generates robust formulation with increased certainty and consistency and this is considered as a distinct benefit where the regulatory issues are important.

(iii)Training aids.

All the system has been used to provide training for both novice and experienced formulator.

(iv)Speed of development.

Reduction in the duration of the formulation process has been reported by many formulators who are using expert system.

(v)Cost saving.

It can be achieved by reducing a development time and more effective utilization of ingredient.

(vi)Freeing the experts.

The implementation of expert system in formulation has inevitable allowed expert formulator to spend more time for innovation.

(vii)Improved the communication.

Expert systems in a company have provided a common platform from which to discuss and manage changes in working practice and to identify those critical areas where the research is required.

(viii)Improvement in formulation.

It provides opportunities to improve the formulation which may be extend product life of existing drug.

Expert Systems in Development of Pharmaceutical Formulation

Expert system may provide an affordable means of capturing knowledge and expertise in a documented form which available to all for free access. Literature survey reveals that following expert systems are become integral part of product development strategy in various dosage forms.  Table 1 summarised list of expert systems used in development of various pharmaceutical formulation.

1. Formulogic System:

It is developed by Logica UK Ltd with reusable software kernel and associated methodology to support the development of generic formulation process. Individual formulation applications are developed using the shell by defining the characteristics of the domain and the corresponding approach of formulation.  Finally, it ends up with the decision making support tool for the formulators that provide the assistance in all aspects.

2. BootsSystem :

This system was developed for skin care formulation. The Formulogic is used as developing tool which known as SOLTAN, which use knowledge captured by interviewing the senior formulator. In addition, the existing information sources such as databases are presented in a frame based semantic network and that can be manipulated by the problem solving domain. Tasks are organized in hierarchy that is build up dynamically depending on the specification at the hand as the problem solving process proceeds.  Knowledge about the formulation is distributed through out the task hierarchy, with strategic knowledge represents towards the top of hierarchy and the tactical knowledge towards the bottom. These developed systems have benefits of cost and better utilization of ingredients.15

3. Cadila System16, 17:

This system was developed by Cadila Laboratory Ltd., Ahmedabad; India for tablet. The PROLOG is used as developing tool based on a 150 prototype of the rules. It is structured in two knowledge bases in a spread sheet format that include the selection of excipients for active ingredients based on their physical, chemical and biological properties and interaction between ingredients. This system is able to identify most compatible excipients for the active ingredient. In this system, the rows represent the property of active ingredient, solubility, and density. Procedural model for the development of Cadila expert system for tablet formulation is shown in figure 3.

4. Galenical Development system 18, 19, 20:

This system was developed by personnel of University of Heidelberg, Germany for aerosol, injection, capsule and tablet. It was recently upgraded using SMARTTALK V running under the window operating systems on a computer. It is design to provide assistance in throughout the development process of formulation. Knowledge basis included all the aspect such as drug and excipients property, medicaments compatibility with added excipients, processing operation, packaging containers and the storage container which gives the reliability and accuracy factors lies in between 0 to 1.

5.  University of London/Capsugel System 21, 22, 23:

This system was developed by  University of London and supported by Daumesnil of Capsugel for hard gelatin capsule. It uses production rule with a decision tree implemented in ‘C’, coupled with user interface through which the user can access both the database and develop new formulation. In this system, all necessary inputs have been design in questionnaire form. It requires the information on physical properties of active ingredient (e.g. dose, particle size, solubility, wettability, bulk density, melting point etc.), compatibility with excipients and the specific manufacturing conditions used by the user.

The systems provides an output package that includes all the necessary information required for processing and filling of powder, capsule size, statistical optimize formulation, specification of excipients, recommended test for quality of product and complete documentation of formulated product.

6. Sanofi system 24 :

It was developed by research division of Sanofi for the formulation of hard gelatin capsules based on specific pre-formulation data of active ingredients. It is based on Formulogic developing tool.

7. Zeneca System 24, 25, 26 :

This system was developed by Astrazeneca Pharmaceuticals Ltd.; UK for tablets, film coated tablet and parentrals. It is based on Formulogic developing tool. The complete system is divided into three stages such as the entry of data, product specification and strategy, identification of initial formulation and formulation optimization.

8. GlaxoSmithKline (GSK) and Astrazeneca (AZ) System 27, 28 :

This was developed bypaternership between GlaxoSmithKline, Astrazeneca and Intelligensys for oral solid dosage form. It is  helpful for the prediction of cracking in film coatings; optimisation of capsule filling; and modeling powder packing to simulate tablet compaction.

• Tablet cracking simulator: It is able to simulate crack propagation in viscoelastic materials. It helps researchers to view how the crack propagates; at what speed and to what degree the cracking occurs across the simulation box. 

• Powder packing simulator: It is useful in solving problems involving particulates and particle packing to help understand the behavior of particulates in materials and able to solve problems associated with the variability in the filling process e.g. estimating volumes, materials handling loads and mixing problems etc. The knowledge gained can underpin aspects of formulation design e.g. capsule design and filling, ratios and distribution parameters.

• Tablet compaction simulator: It is used to simulate compaction of particulates on a large-scale manufacturing of tablets. It is able to address the variability of a multi-particulate system and calculate stresses in the system to generate knowledge as to how the compaction would behave during large-scale manufacture. It can be helpful to avoid mechanical failure (capping, lamination); to optimize process conditions e.g. rate, waste, rejects and scale up at industrial level


The high cost involved in the development of formulation has diverted the pharmaceutical companies to find out alternate strategies. Expert systems are one of the most promising strategies for formulators and IT researchers. This article describes various expert systems used for the development of the formulations. The systems are capable to analyaze and interpret the large amount of complex data and technical details which was required for the development of formulations. The implementation of such system may discourage enthusiastic participation of young scientist in the development of new formulation. This system cannot be substitute of expert formulator but it act as a decision supporting tool which have invaluable benefits to company and society. Considerable intellectual and implementation challenges lay ahead but the potential is required rewards will completely transform how to develop and manufacture pharmaceutical formulation in future with expert system.

Structure of Cadila expert system for tablet

Figure 3:  Structure of Cadila expert system for tablet.

(courtesy: Structure of expert system developed by Rowe for the formulation of tablet)                                   

Sr. No.

Name of Expert System

Name of Company / Institute


Development Tool

Application or Uses of system


Formulogic System

Logica UK Ltd.

Generic formulation

Conventional language

System is used for generic formulation it gives assistance in all the aspect of formulation development.


Boots System

Boots Company



System is use to develop and select other excipients for oil in water, water in oil emulsion, lotion and cream.


Cadila system

Cadila Laboratories (India)



System is use to select optimum concentration of excipients with good compatibility between active ingredients and excipients.


Galenical development system

University of Heidelberg

Aerosols, Tablets, Capsules, Injection


System is use to optimized product.


University of London/Capsugel System

University of London / Capsugel




System is used  for selection of capsule size, quantitative optimization of excipients and the validation of formulation .


Sanofi System

Sanofi Research



System is used to select the excipients for capsule


Zeneca System

Zeneca Pharmaceuticals

Tablets, Parentrals, Film coatings


System is used for the  development of the formulation of tablet parentrals and film coating process


GlaxoSmithKline (GSK) and Astrazeneca (AZ) System

GlaxoSmithKline, Astrazeneca and Intelligensys

Film coatings, capsule filling, tablet compaction


System is used for Film coatings, capsule filling, tablet compaction

Table 1:  List of expert systems used in development of various pharmaceutical formulation.


1.Abadi SSR. Knowledge management in health care: towards ‘Knowledge driven’

Decision– support Services. Int  j. of medical informatics. 2001; 63:5-18.

2.   Rowe RC, Roberts RJ. Expert systems in pharmaceutical product development. Encyclopedia of Pharmaceutical Technology. Marcel–Dekker; 2nd edn,  2002; 1188-1210.

3. Yu LX. Applications of process analytical technology to crystallization processes. Advanced Drug Delivery Reviews. 2004;56: 349-369.

4. Bently P. Product formulation expert system. Intelligent software for product formulation.London, Rowe, R.C. Roberts, R.J. Eds.; Taylor and Francis: 1998;  27-41.

5.  Wood M. Expert system save formulation time. Lab Equipment Digest.  1991;      17 -19.

6. Patridge D, Hussein Ch M. Knowledge based information system. Mc-Graw Hill; New York; 1994.

7. Bradshaw D. The computer learn from Experts Financial times; London; 1989:26.

8.  Turner, J   Manufacturing Intelligence. 1991; 8:12-14.

9.  Rowe RC. Applying neural computing to product formulation. Manufacturing Chemist. 1996; 67: 21-23.

10.  Chunhua Zhao, Ankur jain, Leaelaf Hailemariam, Pradeep Suresh, Pavankumar akkisetty, Girish joglekar, venkat venkatasubramanian, Gintaras V. Reklaitis, Ken Morris, and Prabir Basu. Toward intelligent decision support for pharmaceutical product development. J. Pharmaceutical Innovation. 2006 ; 5: 23 – 35.

11.  Frank J..Knowledge-based assistance for the development of drugs. IEEE   Expert-Intelligent Systems & Their Applications.1997;12: 40-48.

12.  Ramani KV., Patel M. R., Patel S.K. An expert system for drug preformulation in a pharmaceutical company. Interface. 1992; 22: 101-108. 

13.  Rowe, RD. Case-based reasoning–a new approach to tablet formulation. Pharm. Tech. Eur.1999; 11: 36-40.

14.  Rowe RC. An expert system for the formulation of pharmaceutical Tablet. J. Manu. intel. 1993; 55: 1040-1045.

15.  Kashihara T, Yoshioka M. Assessment in Japanese focus group.  Formulation design of oral dosage form: Hashida, M. Ed; Yakagyo-jiho: 1998; 244-253.

16.  Striker H, Haux R, Wetter T, Mann G, Oberhammer L, Flister J, Fuchs S, Schmelmer V. Das.  Galenische Entwicklungs-system Heidelberg. Pharm. Ind.1991; 53: 571-578.

17.Striker H, Fuchs S, Haux R, Rossler R,  Rupprecht B, Schmelmer V, Wiegel S. Das Galenische Entwicklungs-system Heidelberg-Systematische Rezepturentwicklung. Pharm. Ind. 1994; 56: 641-647.

18.Frank J, Rupprecht B, Schmelmer V. Knowledge–based assistance for the development of drug. IEEE expert. 1997; 12 (1): 40-48.

19.Lai S. An expert system for the development of powder filled hard gelatin capsule formulations. Pharm. Res. 1995; 12:  150.

20.Lai S. An expert system to aid the development of capsule formulations. Pharmaceutical Technology Europe.1996;8: 60-68.

21.Newton JM, Podczeck F, Lai S, Daumesnil R. The Design of an Expert system to Aid the development of Capsule Formulations. Formulation Design of Oral Dosage Forms; Hashida, M.Ed.; Yakugyo-Jiho. 1998; 236-244.

22.Bateman SD. The development and validation of a capsule formulation knowledge-based system. Pharm. Technol. 1996;20: 174-184.

23.Rowe RC, Colburn EA. Generating rules for tablet formulation. Pharm. Tech. Int. 2000;12: 24-27.

24.Rowe RC. Film coating formulation using an expert system. Pharm. Tech. Eur.1998;10: 72-82.

25. Rowe RC. Expert system for parentrals development. PDA J. Pharm. Sci. Technol. 1995;49: 257-261.

26.Rowe RC. Expert-systems in solid dosage development. Pharm. Ind. 1993; 55: 1040-1045.



About Authors:

Tejas Thakkar

Mr. Tejas Thakkar is working as Lecturer in Computer department of N. V. Patel College of pure and applied sciences at Vallabh Vidyanagar, Gujarat, India. He holds degree degrees in master in computer application and business administration. His main areas of interest are the expert system and its application in development of various formulations.

Dr. Priti Sajja

Dr. Priti Sajja is currently working as a reader in Computer Department of Sardar Patel University at Vallabh Vidyanagar, Gujarat, India. She earned her Ph. D. Degree in 2000 and  M.Sc. in Computer Science in 1993 from Sardar Patel University. She is member of Computer Society of India (CSI), Special Interest Group on Artificial Intelligence (SIGAI), Third World Organization for Women in Science (TWOWS) Italy,

Dr. Tejal R. Gandhi

Dr. Tejal R. Gandhi is currently working as principal Of Anand   Pharmacy College ANAND. She has 12 years of regular teaching, 2 years as a research scientist experience. She has more than 10 Research publications in international and national pharmaceutical journals. She has also presented papers at various conferences. She earned her MPharm and PhD in Pharmacology

Mrs. Vaishali T.Thakkar is currently working as an Assistant Professor in Pharmaceutics, Anand Pharmacy College ANAND. She has 6 years of teaching Experience She earned her M Pharm in Pharmaceutics and Is actively working toward her PhD at Sardar Patel University

Mrs. Purvi A. Shah is currently working as an Assistant  Professor inPharmaceutical Chemistry , Anand Pharmacy College ANAND. She has 6 years of teaching Experience She earned her M Pharm in Quality assurance and is actively working toward her PhD at Sardar Patel University

Volumes and Issues: 

Similar Entries