Zachary Woods
February 18, 2021
Market demand is shifting at an ever-increasing clip. The world’s most recognizable brands (think Apple in electronics, L’Oreal in skincare or Tesla in automotive) rely on enabling chemical and material technologies. Chemical and material companies are in a race to meet these new market needs. First movers will be rewarded handsomely. However, unlike digitally native industry disruptors like Amazon or Netflix, they don’t have the “data first” culture required to accelerate product development in the near term and create defensible competitive advantages in the long term. While these firms regularly tout significant investment in R&D, often ranging from 5-15% of annual revenues, very little if any Return on Research Capital is traceable[1].
Poor data infrastructure kills the efficiency required to compete today and inhibits AI which will be required to compete tomorrow.
Historically, pundits discussed enterprise data in the context of the five V’s: volume, variety, velocity, validity and value[2]. This framework is great for characterizing data along these dimensions. However, it’s time for a new paradigm that directly aligns with the requirements for running artificial intelligence (AI), enabling the Holy Grail of predictive chemistry. We at Alchemy propose the Five C’s:
Most chemicals and materials companies use a multitude of software tools in their labs to complement paper lab notebooks and Excel. These tools include formulation software, electronic lab notebooks (ELN), laboratory information management systems (LIMS), and many others. Together, they create problematic data silos, each based on different structures, labeling procedures and formatting of data. To compound the problem, many chemists and scientists store Excel data on personal computers and in LAN files. In aggregate, lab data is overlapping, inconsistent, incomplete, disconnected, and generally of low value as it cannot be easily searched, reused, or leveraged by AI.
Most organizations are simply trying to digitize their data and store it securely in a single digital repository. They often do not have the bandwidth (or the underlying infrastructure for that matter) to ensure their data capture meets the Five C’s. And yet, these are the criteria required for any organization to run AI which can dramatically speed product development and reduce testing cycles.
Gartner reports that nearly 60% of organizations don’t measure the annual financial cost of poor quality data[4]. Poor data quality puts organizations on the defensive, resulting in slower innovation while increasing the risk a competitor beats you to market.
To overcome this data challenge, chemicals and materials companies need to address their data quality issues at the source, namely with their chemists as they formulate. With features of ELN and LIMS built right in, Alchemy connects your core systems of record together to offer more context to the researcher while working double duty for the purposes of AI.
The key impact areas that can be addressed with better data management in the lab:
