There’s no doubt about it. AI (Artificial Intelligence) has been the buzzword for 2021 and looks to be an even bigger focus in 2022. But there are also several other trends that will also impact industries and their laboratories across the board through the coming year, and we’ll examine each one and what they mean for laboratories. The underlying theme, however, remains firmly embedded in informatics, the data technology that enables laboratories to manage the increasingly complex nature of the science they perform.
Big pharma players are no exception, with Novartis continuing its data tech partnership with Microsoft. Their AI Innovation Lab, established specifically to facilitate the use of AI throughout the entire company, is looking into these kinds of areas:
> Natural Language Processing (NLP) in areas such as: information extraction, information retrieval, question-answering, summarization, text classification, generative models
> Deep learning models for NLP
> Extensive Python programming
> Machine Learning in areas such as PyTorch, tensor flow, transformer models, BERT, deep learning, neural networks, autoencoders
> Biomedical and pharma data sets and NLP methods
In fact, Novartis professes a major commitment to data technology in a variety of ways on its site.
All of the players in pharma especially realize the major role of data management technology, as noted in Roche’s home site, for example.
And with even greater buy-in to gene therapy across the spectrum, notably immuno-oncology therapy (Roche, many others) and things like ocular therapy (Novartis’ latest acquisition), it’s clear that efficient, accurate and effective FAIR data management will become even more crucial.
Of course, the hangover from 2020/21 and likely to be, to at least some degree, a continuing factor in planning and operations in 2022 is SARS-CoV-2. At this time of writing the predominant variant Delta strain has largely given way to the much more contagious but apparently less deadly Omicron strain. And while its risks may be less substantial, businesses and individuals continue to vaccinate, wear masks and avoid crowds, and testing is actually increasing, since symptoms may not be present, or may be mild. A negative test remains the ticket to many activities, as does proof of vaccination. The net result is that industries are severely constrained by unavailability of the workforce. Has there ever been a more compelling need for automation and remote tech?
TRENDS FOR 2022
1. Increase in Data: Volume and Types
It’s all about the data, let’s face it. Data are ultimately what labs are about. Buy as many expensive instruments as possible and employ tip-top professionals, but if at the end of the day you produce no data you have failed.
Plan to see those data increase in number and expand in variety, as data-heavy areas like genetic sequencing (already gaining more and more momentum), boosted by necessary PCR advances in capacity and throughput because of the pandemic, grow. Add AI and big data and you have a hot bed of activity. This means adopting strategies for building data models to drive data-driven decisions, rigorously capturing as much metadata as possible. Much has been written about “The Lab of the Future,” and “Laboratory 4.0.” The year 2022 may well be the data proliferation point that constitutes the platform for that concept to be realized.
2. Increase In Data Integration and Management
Along with the trend simply toward more and varied data comes the natural necessity to organizing those data and communicating them between data sources. With such a high volume and disparity of data to manage and integrate, standards and strategies become not just important, but crucial. Selection and implementation of those standards will begin to settle and become more, er, standard, and labs will use them as they begin to build a Digital Business Platform (DBP), essentially an ecosystem of specialized modules revolving around a data lake, rather than one monolithic software that is expected to do it all.
AI looks to play an important part in the DBPs that 2022 sees emerge. There is also a trend away from proprietary standards, however entrenched, such as Microsoft, in favor of open source software, giving greater user control and power over their own destiny – rather than being subject to seemingly arbitrary cessation of support, “updates” that force additional purchases and/or reconfiguration of IT architecture, etc.
With LabTwin, we are offering a new technology that helps labs take major steps toward achieving that “Lab of the Future” vision, by providing an innovative hands-free technology for data exchange which can be integrated with any IT system (LIMS, ELN, …) or database, as well as lab instruments.
3. Greater role for AI
As mentioned earlier, AI will increasingly permeate all informatics and automation to one degree or another, helping to future-proof labs through continual adaptation. Newer tech solutions like LabTwin are incorporating it to automate data processing. With the pressure on for scientists to gather more data faster, the application of AI provides a much-needed boost that allows them to concentrate on science.
Lab users can step into the lab of the future by loading protocols into LabTwin, which then provides voice-guidance for execution of protocols, freeing the scientist to conduct processes. In essence it is a digital lab assistant. But LabTwin being smart, it can leverage Natural Language Processing and entities recognition to automatically enrich, structure, interlink and label the data. This data can then be directly exported in a standardized format for further analysis or integrated within a structured data repository.
LabTwin’s exciting capabilities are made possible through careful engineering combined with data science, using an ontological approach that enables ready recognition and handling of technical and other laboratory terminology and processes. The real-time voice interfacing with the lab user incorporates both NLP and AI for outstanding performance even given the wide variety of languages and accents found across the laboratory community.
An example of AI’s growing importance is also reflected in Georgia Tech’s recent government funding: https://www.research.gatech.edu/georgia-tech-awarded-commerce-department-grant-develop-ai-manufacturing-economic-corridor).
4. Increase in Collaborative Processes
Now that there is almost universal acceptance of cloud technology as secure and effective, collaboration has become much more possible. Many labs are already incorporating the benefits of cloud-based technology, and others, especially those doing drug development and other types of research, will begin to leverage this more and more. COVID challenges have led to remote work becoming a proven model that will probably remain to some degree, regardless of the course of the pandemic, if not become much more commonplace. As more and varied data are generated, integrated and managed, the human resources will also expand regardless of geography, to fully leverage the expanded throughput and pace they bring about.
If it seems like data and data management figure predominantly in the outlook for 2022, you’re getting the idea. Yes, there are certainly trends in various areas of the laboratory industry: nanotech, biotech, genetics, robotics…but underlying it all is a definite trend to automation and data management. Old-fashioned lab management and processes still work. But increasingly, manual-based labs will continue to fall behind those who leverage tech to increase their productivity and viability. Last year is now behind us, and unfortunately some labs will also be consigned to memories. Looking toward the future – and more importantly acting accordingly – is fast becoming less of an idle contemplation and more of a matter of survival.