AI: Accelerating innovation in Life Sciences

Digital technologies have already changed the way we research, develop, and manufacture medicines as well as impacting how patients and their physicians decide when and how they receive medical treatment. Yet, despite increasing levels of adoption, life sciences companies continue to lag in digital maturity and are far from realising their full potential to deliver tangible business value.

Artificial Intelligence (AI) in particular is generating a lot of excitement in the sector, particularly generative AI (gen AI), where, according to PWC, 200 use cases have been identified that would deliver ‘material’ benefit to life sciences companies. McKinsey also recently included life sciences as one of the top three industries that could achieve the biggest financial benefits from gen AI, quantifying the gains as 2.6-4.5% of revenue or $60 – $110bn annually. They identified the greatest value potential areas to be within the Commercial and R&D functions, but quantified benefits across the entire value chain. 

Whilst it is likely to take years for the full value of AI to be realised, the industry is making investments and bold moves to build, partner, and/or acquire capabilities and explore use cases. In R&D for example, Novartis & Lilly announced partnerships with DeepMind’s AI drug discovery spin-off Isomorphic Labs. Fiona Marshall, President of Biomedical Research at Novartis, noted that one of the “low-hanging fruit” areas of focus for the collaboration is to reduce the time from target to candidate molecule. Indeed, this is already playing out with Oxford based AI biotech Exscientia who, in partnership with BMS, have progressed their AI designed molecule for Immunology and Inflammation (I&I) to Phase 1 clinical trials in 11 months, a significant improvement on the industry benchmark of around 3-4 years.

Across R&D, there are many other ways gen AI can drive efficiencies and potentially reduce the drug development timeline and costs. Chat GPT is already being adopted by Pfizer and others to automate the laborious process of analysing scientific literature, picking out relevant content and summarising it for scientists to review when researching both drug targets and drug effectiveness. Novartis and Lilly have invested in Paris based AI company YSEOP to streamline clinical trials through automation and AstraZeneca have launched their own health tech business called Evinova, who will leverage multiple digital technologies, including AI and Machine Learning, to streamline clinical trial design and delivery.

In Manufacturing and Supply Chain, AI can drive many operational efficiencies, but the potential goes even further with digital twins. These are being used to create virtual replicas of manufacturing networks and processes as well as complex global supply chains. They can be particularly helpful in scenario modelling of disruptive impacts that could compromise critical supplies of medicines for patients. With so much geopolitical uncertainty impacting the industry, digital twins can leverage multiple disparate data sources to accelerate decision making and contingency planning. 

Commercial teams are exploring many opportunities to leverage AI, from automating routine data tasks to delivering hyper-personalised experiences and customised content to healthcare providers (HCPs) – to ultimately support patient outcomes. The potential for AI to deliver multi-level benefits from life sciences into healthcare systems is unquestionable, but it still needs human oversight to manage the applications securely and ethically, understanding AI’s biases and limitations for patient related applications. 

Real Estate Impacts

A number of pharma companies have developed their own in-house AI R&D hubs in locations rich with digital capabilities. King’s Cross, London, is the location of choice for both GSK and more recently Novo Nordisk, who attributed their decision to locate their AI R&D hub in the Knowledge Quarter to supporting collaboration with the research institutions, big tech, and innovative startups, as well as attracting top talent. As the King’s Cross ecosystem further develops and matures over the next few years, it’s draw for similar industry requirements is only likely to strengthen.

Automation of manufacturing will inevitably have longer-term impacts on both real estate footprints and workforce profiles. McKinsey suggested back in 2021 that if asset utilisation, measured as overall equipment efficiency (OEE), increased from the industry average of 35% to 60%, only 30% of existing manufacturing sites would be required, resulting in a 70% reduction in manufacturing footprint required. We are witnessing that the new, highly digitalised manufacturing sites are smaller, but more scalable. However, for now, this seems to be offset with the increasing demands for medicines and a global shortage of manufacturing capacity for many leading companies – who are therefore expanding, not reducing manufacturing footprints.

For offices, the biggest impacts are likely to be the changing functions, roles, skillset dynamics, and collaborations within them, so to that effect, more emphasis will be placed on reviewing and evolving the workplace design and experience with people at the centre.