ARTICULAR: ARtificial inTelligence for Integrated ICT-enabled pharmaceUticaL mAnufactuRing

Lead Research Organisation: University of Strathclyde
Department Name: Inst of Pharmacy and Biomedical Sci

Abstract

There are considerable challenges around digitalisation in science, engineering and manufacturing in part due to the inherent complexity in the data generated and the challenges in creating useful data sets with the scale required to allow big data approaches to identify patterns, trends and useful knowledge. Whilst other sectors are now realising the power of predictive data analytics; social media platforms, online retailers and advertisers, for example; much of the pharmaceutical manufacturing R&D community struggle with modest, poorly interconnected datasets, which ultimately tend to have short useful lifespans.

A result of poor, under-utilised datasets, is that it is largely impossible to avoid "starting at the beginning" for every new drug that needs to be manufactured, which is very costly with new medicines currently doubling in cost every nine years; $1 billion US Dollars currently "buys" only half a new drug so addressing this issue is key for sustainability of the industry and future medicines supply. This project, ARTICULAR, will seek to develop novel machine learning approaches, a branch of artificial intelligence research, to learn from past and present manufacturing data and create new knowledge that aids in crucial manufacturing decisions. Machine learning approaches have been successfully applied to inform aspects of drug discovery, upstream of pharmaceutical manufacturing, where large genomic and molecule screening datasets provide rich information sources for analysis and training artificial intelligences (AI). They have also shown promise in classifying and predicting outcomes from individual unit operations used in medicines manufacturing, such as crystallisation. For the first time, there is an opportunity to use AI approaches to learn from the data and models from across multiple previous development and manufacturing efforts and then address the most commonly encountered problems when manufacturing new pharmaceutical products, which are knowing: (1) the processes and operations to employ; (2) the sensors and measurements to deploy to optimally deliver the product; and (3) the potential process upsets and their future impact on the quality of the medicine manufactured.

All of these data and the AI "learning" will be made available via bespoke, personalisable AR and VR interfaces incorporating gesture and voice inputs alongside more traditional approaches such as dashboards. These immersive interfaces will facilitate pharmaceutical manufacturing process design, and visualisation of the complex data being captured and analysed in real-time. Detailed, interactive 3D visualisations of drug forms, products, equipment and manufacturing processes and their associated data will be created which provide intuitive access across the length scales of transformations involved from the drug molecule to final drug product. This will be unique tool, allowing the user to see their work and engage with their data in the context of upstream and downstream processes and performance data. Virtual and Augmented Reality technologies will be used in the lab/plant environment to visualise live data streams for process equipment as the next step in digitalisation. These advanced visualisation tools will add data rich, interactive visualisation to aid researchers in their work, allowing them to focus on the meaning of results and freeing them from menial manual data curation steps.

Planned Impact

The ARTICULAR team engage and collaborate with a wide range of research beneficiaries: leading pharma companies (e.g. AZ, GSK, Bayer, Pfizer, Lilly); companies in the process industries (e.g. Mars, AB Sugar, Syngenta); technology and supply chain companies (e.g. Siemens). ARTICULAR team members are actively involved in the innovation landscape with collaborators including; HVM Catapult (CPI); the National Formulation Centre; medical charities (e.g. CRUK), Medicines Manufacturing Industry Partnership (MMIP) of the UK pharma manufacturing industry; International academic partners (Rutgers, Graz, Singapore) through the newly formed International Institute of Advanced Pharmaceutical Manufacturing. This strong, active and influential network of collaborators, stakeholders and business leaders maximises the potential to achieve the project Aims with associated impact and benefits including:

Economic: The project will contribute to the UK economy by increasing the competitiveness of pharmaceutical and technology partner companies with whom we will co-develop our commercialisation and exploitation plans. Companies will have access to the modular AI-driven ICT tools and applications that emerge from the project. This will increase competitiveness by accelerating process development using minimal amounts of API, improve quality control and minimize waste. The programme will also deliver cost effectiveness via novel adaptive control using machine learning to interpret data. In addition to implementation across CMACs £34m physical hub, establishing a world leading digital lab capability, our connectivity with the HVM (CPI) and Digital Catapults, Diamond and NPL will form critical links with the wider innovation system helping drive industrial translation where opportunities for existing and startup company benefits to be realised are identified

Societal: A key societal impact is to improve the accessibility and affordability of new and existing drug products and the seed at which they can be brought to market. ARTICULAR will do this by lowering development and production costs through intelligent systems, improving equipment utilisation, reducing energy consumption, waste and reject reduction, improved efficiency of material consumption and reduced environmental impacts. With machine learning, automated, intelligent control has the potential to enable intelligent decision making. These technology developments will have a major international impact, with significant potential for distributed manufacturing: small modular, reconfigurable, manufacturing capability with automated, intelligent control has the potential to enable manufacture close to the point of need. Furthermore, opportunities for anti-malarial or anti-retroviral products being made at low cost in developing economies for local distribution to patients will improve access to safe and effective healthcare in these poorer regions of the world.

Academic: The world-leading academic partner combination incorporating data analytics, control, visualisation, pharmaceutical sciences and chemical engineering will contribute within and across disciplines producing significant advances in theory and understanding directed towards innovative methodologies, equipment and techniques. A key output will be delivering a highly skilled talent pipeline of trained researchers capable of transferring the generated research and knowledge into UK based multinationals and SMEs to enhance the knowledge economy and revolutionise capability in AI for medicines manufacturing. As a further route to impact, we will provide new opportunities for them to obtain academic and industrial experience via researcher exchanges. CMAC is a recognised academic leader and its unique role was outlined in the Medicines Manufacturing Industry Partnership 'Manufacturing Vision for UK Pharma'. Through our active dissemination and outreach programmes we will maximise the lasting legacy of this research.

Publications

10 25 50
 
Title Process twins 
Description 3D models of medicine manufacturing process equipment. Assets include: Laboratory environment COBR crystalliser Hot-melt extruder Lab-scale reactor Automated mixer Filtration equipment Tablet press 
Type Of Art Artefact (including digital) 
Year Produced 2019 
Impact Work here has attracted the attention of companies within the UK and has led to a KTP application to further developments (awaiting outcome). 
 
Description The use of Artificial Intelligence to provide insight from data is showing particular promise in certain areas of medicines manufacturing we are currently developing.

In particular:

- Imaging of drug particles to better describe their size and shape and to convey this complex information to humans is possible through machine learning (ML) approaches. This rich information can also be used to control and optimise the medicines being manufactured to ensure better products for patients

- When developing new medicines it is helpful to be able to predict the outcome from processes, such as crystallisation, in which drug materials are made. We have generated AI/ML tools which are able to tell researchers about the likely form a drug crystal will take (its constituent molecules) and the likely shape or morphology. Using these tool, it is possible to inform design choices to ensure that materials are simpler to handle and will perform well in patients

- We have developed new virtual environments of labs and processing equipment which we intend to connect to data generated by sensors on the manufacturing processes themselves. This will provide novel ways for workers to engage with their data and surroundings and potentially allow for real-time cross site operations and collaborations

- We are currently beginning studies of the effectiveness of VR/AR as data visualisation tools in the development of digital twins of medicine manufacturing equipment and processes. The outcomes of these studies will potentially demystify some of the hype around VR/AR technologies and support further developments in the field.
Exploitation Route Broadly, many of our digital twin demonstrators to date rely on commodity VR hardware which will currently be available in some homes. Over the next couple of years, we will likely see wider penetration of these technologies which would then allow us to make developments available to the public. In particular, undergraduate and postgraduate cohorts are likely to benefit from additional insight into the manufacture of medicines being able to immerse themselves within a lab or manufacturing plant from home and to interact with processes and equipment to learn more about what they do and how they operate. In a world where patient compliance with regard to taking medicines is a major issue, education on the complexities and costs of producing these treatments could potentially be informative and reduce costly wastage.

We are currently exploring the exploitation of ARTICULAR demonstrators being developing with the EPSRC Future Manufacturing Research Hub for Continuous Manufacturing and Advanced Crystallisation (CMAC) and partners. These include seven of the largest pharmaceutical companies some of which are actively demoing our VR environments on their sites to raise the profile of Industry 4.0 technologies internally and with stakeholders.
Sectors Chemicals,Digital/Communication/Information Technologies (including Software),Healthcare,Pharmaceuticals and Medical Biotechnology

 
Title CrystalEyes 
Description An image analysis and Machine Learning (ML) model which is able to accurately model and classify particles by their shape from segmented images. Morphometric analysis is used to describe the shape in great detail and then these data can be used to assess large numbers of images to describe crystal shape populations. These shape descriptors can can also be correlated to processing steps to provide more control over particle shape during manufacture. 
Type Of Material Computer model/algorithm 
Year Produced 2019 
Provided To Others? No  
Impact In-house tool at present but with intention to make available to wider research community. 
 
Title Tablet press - data management and analysis 
Description This code is useful for extracting detailed information derived from the sensors in a tablet press providing operators with visualization tools for tablet-to-tablet, batch-to-batch variation and that, with further AI developments, should provide opportunities for process fault detection which would include events such as material sticking. 
Type Of Material Data analysis technique 
Year Produced 2019 
Provided To Others? No  
Impact We are currently developing these codes and will investigate their utility within the EPSRC CMAC programme and potentially in the Medicines Manufacturing Innovation Centre Grand Challenge 1. 
 
Description Digital Twin development with Perceptive Engineering Ltd 
Organisation Perceptive Engineering Ltd
Country United Kingdom 
Sector Private 
PI Contribution We are currently working with Perceptive Engineering Ltd to deploy some of our process monitoring and control approaches on one of their demonstrator platforms. As a part of this work we have developed a virtual reality digital twin of their mixer platform and integration code to pass data from their PharmaMV software to the VR interface.
Collaborator Contribution They have provided access to their databases and software that enables us to pass data to our VR front-end and have provided their mixing demonstrator hardware for us to model. By working together, we have created a simple demonstrator of AI controlled processes with VR digital twin.
Impact This new demo is small, mobile and being demonstrated at events to give a broader audience exposure to Industry 4.0, the use of AI in process monitoring and control and the ARTICULAR project. It has created a new interface to the PharmaMV software which will be exploited across other CMAC platforms in the coming year. We are also currently preparing a 2-year KTP application to provide additional resource for working directly with Perceptive Engineering.
Start Year 2019
 
Description Keynote 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Chapman, Paul (2019) Invited Keynote for the 5th Art and Science International Exhibition Symposium, China. In: 5th International Art and Science Exhibition and Symposium, 1-3 November 2019, Beijing, China
Year(s) Of Engagement Activity 2019
 
Description Keynote 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact digital | visual | cultural 3 - models, volume, and digital 3D visualisation
Chapman, Paul digital | visual | cultural 3 - models, volume, and digital 3D visualisation. In: digital | visual | cultural 3 - models, volume, and digital 3D visualisation, 17/18 June.2019, Oxford University.
Year(s) Of Engagement Activity 2019
 
Description National Sarcoma Conference 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Third sector organisations
Results and Impact Presentation of ARTICULAR research - The British Sarcoma Group, National Sarcoma Conference, Wednesday 26th and Thursday 27th February 2020, Hilton Hotel, Glasgow
Year(s) Of Engagement Activity 2020
 
Description The Future of Medical Visualisation 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The Future of Medical Visualisation - Paul Chapman, School of Simulation and Visualisation, Glasgow School of Art
Year(s) Of Engagement Activity 2020