Artificial intelligence lab design

Decoding Tech Trends for the Lab

It’s becoming very common to see “trend” lists populate our social media and news streams. Experts across all sorts of industries try to predict exactly what big developments will define the next year, the next five years, the next decade, etc. What will make headlines? What will disrupt the market in a way that no one anticipated? Life sciences have seen a number of big tech trends re-occur on these lists several times over the last few years. When it comes to laboratory design, it’s important to go beyond awareness of the trend and translate how it could impact the space, equipment, operational flows, etc. In this article, I take a look at two trends that have been mainstays on these lists for the past few years — predictive analytics and artificial intelligence — and share tangible ways they could impact the way we program labs and other research spaces.

Predictive analytics offer promise for speed to market.

The Promise of Prediction

The process of drug discovery is long, drawn-out and expensive, taking anywhere from 10 to 15 years and costing, on average, $1-2 billion1 to develop a scalable drug candidate.

In an effort to speed-up this process, corporations have looked to technology as the magic bullet for improved targeted drug development. In the 1990s and 2000s, high-throughput screening (HTS) was seen as the technology that would revolutionize, expedite and drive the future of drug discovery. The big idea was that by quickly screening a large number of compounds against a given target, the process of predicting a drug target candidate’s viability could be accelerated. This high-risk process led to many exciting targets in the early discovery phase, however, the high rate of false positives coupled with the quality of the results led to a greater number of the novel compounds not being viable targets to pursue. HTS is no longer viewed as the technological panacea that is going to revolutionize the industry, but instead, is viewed as a valued tool that can supplement and inform the hunt for effective drug candidates.

In more recent years, attention has turned towards “Big Data” and predictive analytics as a potentially more feasible way of expediting drug discovery. Predictive analytics can help take data and lessons learned from both failed and successful drug development efforts to establish connections and aid researchers in making more informed, smarter decisions up front. This ultimately can lead to more viable targets and more successful trials. Already, data mining of available biomedical data has led to a significant increase in target identification. Many pharmaceutical and biotech companies are getting better at using predictive analytics as a way to be “smarter” about identifying key usage patterns and potential treatments that are targeted to a specific set of patients or conditions. Using predictive analytics as a way to evaluate and extend the patent life of an existing drug product for different disease or therapeutic use has also become an area of focus in R&D.

Executing the Analytics

Real-world evidence emerges as a powerful data source.

In recent years, data mining has expanded beyond data from clinical trials to include “real-world evidence” from sources such as medical health records, insurance claims, fitness trackers, health monitors and other wearable technologies. The data coming from these sources is increasingly more reliable. For example, AliveCor’s® ECG KardiaMobile, released in 2016, is the first FDA approved mobile app to provide users with a medically-reputable way to monitor their heart themselves. They have since released an Apple watch band and Apple has released its own FDA-approved technology.

With so many different sources of data, there are serious questions by some if the pharmaceutical industry will ever get to the point of optimizing “Big Data” to drive predictive analytics. Some question whether predictive analytics will go the way of high-throughput screening and recede back to the role of “tool” rather than being the predicted game-changer. However, in the past couple years, several companies have continued to forge ahead, betting big that if the data can be appropriately harnessed, analyzed, and applied, the payoff could be massive. And success cases are starting to emerge with more targeted personalized medicines being developed and tools with the ability to predict study risks and provide preventative and maintenance measures being deployed.  

Artificial intelligence and machine learning offer potential at all stages of drug discovery.

The Potential of AI

Aiding the promise of predictive analytics is the incorporation of intelligent enterprise technologies, including artificial intelligence and machine learning. At its core, the power of AI for the pharmaceutical industry lies in its ability to mine and analyze enormous sets of raw data, such as those generated through R&D and clinical trials. Benefits from deploying AI tools such as informatics and data analytics platforms include their ability to make the critical connections and inform research teams of trends seen in trial audit reports and simplify adverse event reporting for practitioners.

The potential of AI-based tools is now being explored at all stages of drug discovery and development — from research data mining and assisting in target identification and validation, to predicting their properties and risks associated with the identification of novel lead compounds and drug candidates. The use of AI-based software can assist in planning chemical synthesis to obtain compounds of interest. AI is also applied to planning pre-clinical and clinical trials and analyzing biomedical and clinical data.

Lab Impact: Increased collaboration of diverse teams from around the globe.

So….What Does This Mean for the Lab of the Future?

The quest to integrate the physical tools that support staff focused on tasks related to data mining and the development of predicative analytic models will have a notable impact on the future of laboratory design. As the process of research discovery becomes more encompassing and complex, there will be a greater need for the collaboration of diverse teams from around the globe. The need to collaborate and share large sets of data virtually will significantly influence physical space requirements of laboratory facilities.   


Lab Impact: More space for dry labs, computational space, and virtually immersive environments.

Building on the trend we have been seeing for years, the ratio of wet lab to dry lab spaces will continue to migrate toward increased dry lab and computational space. Facility space programs will need to account for an increased percentage of technologically rich, smart spaces. These computational environments will drive the need for greater access to reliable, high-speed data infrastructure to allow for super and cloud computing necessary to facilitate and improve collaboration of virtual teams, enabling sharing of large data sets of predictive models. Building programs will also need to incorporate C.A.V.E. or Immersive Environments Data Visualization environments for evolving tools and techniques that are necessary for multi-dimensional data analysisDesigning building infrastructure sophisticated enough to meet current demands, while flexible enough to accommodate a rapidly changing technology in the future will be key.

Lab Impact: Significant growth in data storage needs.

Additionally, the exponential growth in data will translate to significant growth in data storage. Currently, many pharma companies contract with outside vendors for their AI and data analysis capabilities. However, if use of AI and high performance computing continues trending upward, some companies may begin to explore the ROI benefits of bringing these capabilities in-house. Adopting this approach will drive greater demand for finding the right data storage solutions. Bringing high-performance computing in-house will necessitate an investment in added space and infrastructure to handle the large cooling and electrical loads while providing proper redundancy. However, these systems may also provide opportunity to reclaim and reuse waste heat. Further, using cloud-based systems will require proper vetting to ensure the ability to protect personal and propriety data from cyber vulnerability.   

Moving Forward

It can be hard to accurately predict the impact of predictive analytics and artificial intelligence on the pharmaceutical industry in these relatively early stages — there is still so much we don’t know. However, they are important factors to consider. From a facilities and capital planning perspective, it’s important to bring the right people to the table for discussions about what the future might entail from a tech perspective.  

The impact of technology on our society has been nothing short of extraordinary — and the continued impact on healthcare and pharmaceuticals looks very bright. It’s exciting to think about a future with a faster, less expensive drug discovery process that gives expedited access to life-saving vaccines and medicines — a future that also includes continued improvement and innovation of personal monitoring devices that help us take better care of ourselves.

1DiMasi, J. A., Grabowski, H. G., & Hansen, R. W. (2016). Innovation in the pharmaceutical industry: new estimates of R&D costs. Journal of Health Economics, 47, 20-33. + Prasad, V., & Mailankody, S. (2017). Research and development spending to bring a single cancer drug to market and revenues after approval. JAMA internal medicine, 177(11), 1569-1575.