Within life sciences research is an innate conflict between discovery and delivery; discovery requires a comprehensive, meticulous approach to detail and a significant amount of time to achieve true pharma, biotech, or life science advances. Then, those discoveries must be delivered to the market as quickly as possible.
Labs still depending on manual processes are going to lose their competitive edge to the labs embracing full digitization, from automation and AI to advanced data management and robotics.
Leveraging digital transformation to build a future-ready research environment increases efficiency and accuracy, accelerating innovation to meet the demands of both discovery and delivery.
We surveyed more than 150 lab professionals about their labs’ embrace of technology – or lack thereof. Surprisingly, AI wasn’t the top priority. Automation, an “old” technology, came out on top. Furthermore, inventory management – from ensuring that simple reagents are on hand to tracking patient samples – is still a major focus, even though solutions have been available for years, from small systems to enormous automated systems that combine liquid handlers, robotic arms, and more.
Despite these other areas of priority, AI is still top of mind, with 45% of survey respondents expecting to implement AI within the next 18 months.
Big pharma labs, with larger budgets and a greater bottom-line focus, are much further ahead in digitization than startups. Even though startups are expected to be “more agile,” their “seat-of-the-pants” operations don’t often have available capital to invest in preparing for long-term growth.
Recently, SelectScience invited our very own Devin Donnelly (Cenevo) and a full team of experts: Tobias Gafafer (Hamilton Storage), Michael Hopkins (Highres Biosolutions), Jason Meredith (Tecan Switzerland), and Cory Tiller (SPT Labtech) to participate in a webinar about the survey findings and offer thoughts on how to accelerate the digitization process and prepare labs for the adoption of AI.
Automation, like all journeys, starts with a single step. Most critically, automation and all aspects of digitalization need a champion to initiate and manage the process, whether the lab manager or a detail-oriented person who already has enough respect from within and without the lab to be seen as a guide not a threat. The innovation leader needs to focus on the scientists’ actual needs.
While committees sometimes add to the workload, it’s helpful to gather key personnel to visualize where they want the lab to be within the next three to five years. The champion can then build a step-by-step plan for implementation, with the understanding that the technology necessary to get there will be evolving quickly.
Small wins accelerate buy-ins for the greater changes to come. Starting small ensures that a practical system gets built versus creating a comprehensive system from the beginning that looks pretty, costs enormous sums of money, and does little or nothing practical to accelerate discovery.
Labs should start with low-cost tools such as adopting barcodes for samples and supplies and replacing old equipment with more automation-friendly labware. Ensuring full traceability is key. These initial stages are the foundation for the longer-term implementation. They create a smoother process for lab personnel versus trying to automate everything all at once – which, of course, causes significant confusion, usually takes more time than expected, and takes a major chunk of budget upfront.
Hardware automation makes it simpler to screen and manage compounds; software automation supports the lead optimization process to speed drug discovery. No matter what’s chosen, it needs to be scalable to continue to support lab requirements.
One caveat about increasing automation – automation itself creates additional data, which then must be ingested into electronic lab notebooks (ELN) or lab information management systems (LIMS) so it can be used for analyses. Data management should be approached as a step-by-step process as well.
Migration and data transformation from older systems to newer ones that integrate AI models are critical. Statistical process control is also important. Data standardization and integrating that data from a variety of platforms into one central data structure is critical for accurate analyses. Labs should also maintain their focus on FAIR data principles to improve practices and stay in regulatory compliance.
Once the data has been “managed,” AI and ML algorithms can be used in a variety of ways. Large language models can build automation scripts and develop and optimize assays or detect errors, among other tasks. Of course, human insights are critical to the AI process, as the scientists must identify which AI ideas are valuable and which are just noise or gimmicks.
AI can also assist labs with hardware reliability and longevity by analyzing usage to provide prompts for preventative maintenance and making it easier for failure mode recovery. It can also streamline lab efficiency with automated ordering and managing equipment scheduling.
No matter what vendors say about their solutions, no single supplier delivers the holy grail of digital transformation. We offer a comprehensive sample management, ELN, and LIMS solution. Our platform complements and integrates into existing lab technologies, so labs can customize their own end-to-end systems so they can benefit from and don’t suffer because of their mixed informatics environment. If a solution comes with open and extensible APIs, it can be connected to almost any other lab platform.
Devin provided a key insight for streamlining the process of digitalization: "Focus on the bottlenecks that preside over your workflows, be they hardware or software oriented. Then think about it before you make those decisions, do I suddenly create an upstream impact on another bottleneck or a downstream one? You don't want to be moving your pain points around the lab."
Missed the live event? Catch up on the panel discussion on demand now.