Zifo's Survey Highlights the Need for Data Readiness in Biopharma AI Adoption

Zifo's Global Survey on AI in Biopharma

A recent survey from Zifo Technologies sheds light on the landscape of artificial intelligence (AI) and machine learning (ML) within the biopharma sector. The survey captures responses from scientists and informaticians across over 30 science-driven organizations, revealing a dual narrative of excitement and caution regarding AI’s role in research and development. While there's a clear momentum towards increased AI investment, there are significant hurdles in data readiness that could impede progress.

Key Findings from the Survey

The survey indicates that nearly two-thirds of organizations are investing in AI and ML technologies across their operations. However, a mere 32% of respondents feel confident in their company's capability to effectively leverage their scientific data for AI applications. This discrepancy raises pertinent questions about overall data readiness and the practical applicability of AI within these organizations.

One of the most notable insights is the persistent struggle with data accessibility. An alarming 70% of participants reported difficulties in accessing the data required for AI projects, shedding light on the widespread issues related to data silos and a lack of integration. The report pointed out the absence of standardized data storage and metadata practices as significant barriers to progress, underlining the essential need for harmonizing data resources.

As the survey highlights, major obstacles include interoperability challenges and gaps in automation. Nearly half of the organizations indicated encountering difficulties in integrating data from laboratory instruments, complicated by outdated infrastructure and inconsistent standards. While the automation of data capture is increasing, about 26% of organizations still rely heavily on manual processes, and around 10% have no automation at all.

Infrastructure Challenges

Further compounding these issues, the survey revealed that many existing data management solutions are ill-equipped for the demands of High-Performance Computing (HPC) environments. Even though initial data capturing and final storage processes are automated, the intermediate stages of data processing remain poorly supported in most Electronic Lab Notebooks (ELNs).

Paul Denny-Gouldson, Zifo's Chief Scientific Officer, emphasized the critical importance of efficient data management for ensuring data can be reused and retrieved efficiently. This underlying structure is vital for supporting FAIR principles—Findable, Accessible, Interoperable, and Reusable data.

Current Trends and Future Directions

Despite encountering these obstacles, AI adoption across research and development areas is on the rise, with 39% of organizations reporting moderate adoption and another 26% indicating minimal uptake. The areas of research and development represent the strongest focus for AI applications, with clinical applications and manufacturing showing an increasing interest as well. Interestingly, organizations are opting for targeted, incremental AI implementations rather than large-scale, disruptive changes in their operations.

Denny-Gouldson notes that the true value of AI lies in its integration into everyday workflows across various value chains. Key challenges that need addressing include data quality issues, privacy concerns, and difficulties in integrating instruments in lab settings. Organizations are increasingly motivated by the urge to protect their proprietary data, leading many to develop in-house AI solutions to maintain control.

When asked about the prospective advantages of AI, survey respondents highlighted accelerated discoveries, enhanced efficiency, cost savings, and deeper scientific insights as primary benefits. While improved patient outcomes remain the ultimate goal, it appears that immediate, tangible benefits are being prioritized. These intermediate achievements—faster research cycles, streamlined processes, and enriched data insights—are seen as pivotal stepping stones toward enhanced products and better overall outcomes for patients and consumers.

Conclusion: A Call for Data Standardization

Looking ahead, Zifo underscores the necessity for data standardization and seamless exchange among different laboratory instruments for science-driven industries. The organization posits that we are entering an era characterized by 'The Age of Data Management,' which will eventually pave the way for 'The Age of AI.' Those companies that invest effectively in robust data infrastructures, foster cross-functional collaboration, and focus on targeted AI applications are more likely to transform their data into meaningful discoveries, thereby turning innovation into impactful solutions for the industry.