Scroll to top

Enabling a Pharmaceutical R&D Pipeline 

Decision Lab Case Study with GSK


How can computer modelling help us to solve a key need of the pharmaceutical industry – to improve the “backbone” of their business: their R&D pipeline? 


Our challenge was providing the user with a great deal of flexibility in easily defining the problem in the tool, and having it give transparent explanations of the results it produced. 


A transparent and easy-to-use model of an R&D pipeline that simulates the development and maturing of future products, against commercial demand profiles for them to provide insight into progression strategies. 

To answer this tough question, we built a model that has become a real game-changer for GSK

The case study of how Decision Lab UK’s expertise in simulation and the direction from GSK’s team on “Front End Engineering and Design” created a decision support solution for long-term global network chain investments.  

Looking for new ways to optimise an R&D pipeline 

The pharmaceutical industry’s lifeblood is R&D. A major drug company’s success almost completely depends on discovering and developing new medicines, and their capital investment reflects this. Pharmaceutical R&D spending is huge – the second largest of all industries – and it is expected to grow even more. This puts significant pressure on the industry to optimise R&D. However, traditional R&D approaches across the industry yield a low success rate and cost billions, which in turn impacts the return on investment that goes into drug discovery. 

Can new approaches that use exploratory computer modelling help to identify efficiencies and improve the R&D process by predicting the pipeline of drugs in an effective way?  

Imagining the future through the simulation of the chemical assets 

Using agent-based modelling techniques, we developed an advanced simulation environment that relies on information about the chemical assets. The model accounts for assets that currently exist in GSK’s R&D pipeline and predicts future compounds with varying attributes. We have provided an extremely adaptable interface, which allows GSK to tailor model runs to the specific attributes of interest and create new modelled compounds. The user can run a range of scenarios to generate a compelling business case for potential R&D strategies. This includes different progression strategies, such as fast-tracked and accelerated, and their impact on cost and ROI. It handles uncertainty throughout the multiple stages and also in consumer demand for compounds.  

The need for flexibility 

We wanted to develop a way that allowed users to easily and flexibly configure the attributes of the entities or assets within the model. GSK wanted to be able to do this for any number of attributes for any number of assets. Providing this level of flexibility within a simulation and a simple way for the user to specify them was a particular challenge. 

Transparency in results 

The simulation is providing results to inform important investment decisions. For this reason, we did not want to develop a black-box model where there was no visibility of the simulation logic and the outputs could not be easily interpreted. To better support a transparent decision-making process, we created a user-friendly interface that allows users to visualise the progression of products across the pipeline and understand the business logic that has been embedded in each stage of the process.  

The future 

Key concepts of the model are not restricted to the pharmaceutical industry. A similar approach could be applied to any industry with an R&D pipeline. For instance, in the FMCG (fast-moving consumer goods) industry, many iterations of a product are tested to see how effectively they meet consumer needs – the likelihood of a product progressing from the concept stage through to market launch could be simulated with our modelling solution.