Nick Lynch, consultant at The Pistoia Alliance discusses how technology, including artificial intelligence (AI), can help automate the lab and accelerate outcomes.
Automation
Our world has undergone a dramatic shift in the past few decades as technology has developed and digitisation continues apace. However, for the past 20 years, the lab has looked largely the same, even though in that time, how scientists and researchers work and live has changed dramatically. Now, life science companies are seeking ways to modernise research environments and build the ‘Lab of The Future’ (LoTF). Given advances made in recent years, it’s natural that technology that can automate lab functions – including Artificial Intelligence (AI) – will play an important role in this modernisation drive.
There is no doubt that the potential of AI and automation is huge, and its possible applications in life sciences could help to accelerate research outcomes. It is popular, too, with our research finding that more than two thirds (69%) of life science professionals are already using AI in their work. But, realising its potential will also rely on the whole industry to work together and overcome the hurdles to widespread implementation and adoption.
Incubating a new approach to technology
Automation in the LoTF will fundamentally influence how scientists undertake their daily tasks. One of the most tangible ways will be by directly impacting the design and use of physical tools and instruments in the lab. This includes the introduction of robots to automate experiments, which is already relatively commonplace in today’s labs, to more advanced uses such as ‘smart’ machines that can ‘talk’ to users and advise how to handle equipment correctly or offer instructions on how to use an instrument. This will speed up research and ensure the accuracy of experiments, as well as reducing the risk of human error and improving safety.
Eventually, as advances in virtual ‘assistant’ technology continue – such as Amazon’s Alexa and Apple’s Siri – researchers in the lab will be able to speak directly to their surroundings to request information, for instance, before conducting an experiment. We are now seeing these innovations come closer to fruition, taking the automated lab to a new level. Students at the University of Southampton recently won The Pistoia Alliance’s President’s Series Hackathon by integrating data from our Chemical Safety Library into Amazon’s Alexa – giving researchers the ability to access chemical safety data by simply speaking out loud.
Automation and AI also has a role to play for scientists before they even get to the bench – the majority (46%) of AI projects in life sciences currently take place in early discovery or preclinical research phases. Automation here can speed up initial research, with AI platforms able to accelerate data retrieval by automating searching and ‘reading’ literature or comparing biomedical images in a matter of seconds. Scientists can also ‘fail faster’ thanks to automation, modelling multiple scenarios before undertaking work – making research more efficient, cost effective and accurate.
Successful automation requires better collaboration
These developments and the promise they hold for the future of life sciences are exciting; technology is sure to advance even further and the LoTF will take a very different shape than at present. Ultimately, automation will be used to augment researchers, allowing them to reduce the number of manual processes they undertake in the laboratory, focusing instead on the tasks that can’t be done by machines. To reach a point where everyone can benefit though, the life science industry will have to collaborate to overcome the main challenge to the successful implementation of AI and other automation – data. Success from AI is reliant on being able to access high quality data, which is a problem for many; access to data (24%) and data quality (26%) were two of the biggest barriers to AI projects cited in our research.
Introducing industry-wide data standards will go a considerable way to solving this problem, ensuring shared data is useable by all, and the technology can be integrated with existing systems and data sets. A key factor in this will be following guiding principles – such as FAIR (Findable, Accessible, Interoperable, and Reusable) – that aim to facilitate data exchange in increasingly computational research environments. Taking steps to standardise formats will also be important for other aspects of the LoTF that support automation, such as incorporating data gathered from devices and sensors connected to the Internet of Things (IoT), which will come in a variety of different formats and structures. Moreover, successful automation requires skills in both technology and science, and pooling resources to share knowledge is an essential step to improve outcomes.
Beyond data, there are other challenges to contend with that will also require greater teamwork across disciplines. For example, resolving issues with integrating legacy instruments to an automated lab workflow will need collective effort from all stakeholders, including life science companies, software providers, and instrument manufacturers. To provide such a forum for knowledge sharing and bringing together interested parties, The Pistoia Alliance launched the Centre of Excellence (CoE) in Artificial Intelligence and Machine Learning (AI/ML). The CoE is open to all, including non-members, and through events and further research, will help researchers understand how AI and automation can augment their work. Ultimately, patients’ lives depend on making breakthroughs. We encourage anyone interested in AI to get involved in the CoE.