Lab Manager | Run Your Lab Like a Business

Product Focus

two male scientists using tablet in lab

Removing Data Silos with Next-Generation ELNs

Taking chemical R&D from data silos to data-driven success with next-generation informatics

Mary Donlan, PhD,David Gosalvez, PhD

Today’s chemical teams are under increasing pressure to accelerate productivity and speed time to market by better identifying molecules and formulations that will perform best and deliver high product yields. At the same time, they need to look for ways to lower risks, costs, and resource pressures—all while keeping quality high and working to become more sustainable. 

Better leveraging of testing data to help pinpoint successful formulations can serve as a lynchpin in helping meet these challenges. However, most testing data generated in today’s chemical labs is highly complex, manually gathered, and often buried in silos. This makes it difficult and time consuming to quickly and easily find, visualize, reproduce, and share the most valuable data to make more informed, timely decisions and do predictive analysis. This type of environment creates data swamps where information exists but is buried in the weeds and hard to locate when and where you need it. 

Going from data swamps to actionable insights via the cloud

Data swamps are created in chemical labs (or any type of lab) when R&D teams rely on paper notebooks and manual data entry into spreadsheets on individual computers to form the backbone of their data management approach. 

Even when experimental analyses or instrument output data is captured digitally, it can have limited power if siloed. In fact, it has been observed that up to 80 percent of data—even digital data—can become untraceable if it is stored on individual computers.

Within that “lost” data can live great innovation. The key is to get data into a secure but more sharable, flexible, and analysis-rich platform so you can find and use the gems you have.

Digital tools help chemical researchers and R&D teams spur new ideas, document experiments for effective search and reproducibility, do secure sharing, and make decisions around prime development candidates.
Figure 1: Digital tools help chemical researchers and R&D teams spur new ideas, document experiments for effective search and reproducibility, do secure sharing, and make decisions around prime development candidates.

 Secure, cloud-based SaaS informatics ecosystems are geared toward doing just this. Secure, cloud-based SaaS informatics ecosystems are geared toward doing just this—streamlining data capture, search, analysis, visualization, and sharing to find and leverage the critical testing data needed to make more informed formulation decisions. 

SaaS informatics can also loop in materials tracking and manufacturing data to create 360-degree data visibility and holistic decision-making for chemical labs. 

But where should you start? The following are some key elements and best practices to consider when moving or planning to move to SaaS.

Cloud-based ELNs and digital chemical drawing tools set the stage

Whether a lab is working with polymers, electronics, CASE chemicals, or water treatment, scientists and R&D teams need  to cope with and leverage data points from all sorts of formats, sources, instruments, and publications. Team members need to be able to optimize their make-test-decide workflows.

Using a full-function, cloud-native electronic notebook (ELN) with a digital chemical drawing tool is a great starting point to meet these key needs.

Employing innovative digital chemical drawing tools, chemists can more easily create, rotate, and share 3D renderings of their molecules.

With ELNs, scientists can centrally capture, organize, document, and reuse relevant experimental data securely and create custom templates for common experiments, saving time and improving reproducibility. 

They can also conduct Google-like searches to zero in on useful published and historical experiment data created by themselves and others. This helps provide context and trending information that can speed hypotheses formulation and shorten the planning stage of design of experiments.

Researchers can also create and manage custom lists of molecules and formulations based on user-defined properties. For example, those developing new synthetic dyes can add ingredients with certain properties, such as pigmentation, to specific lists. They can then quickly review the collection of materials when looking to design a dye with specific pigmentation properties. 

Next-generation ELNs can also support image or specialty file manipulation without needing to switch programs and feature an automated inventory management component that allows users to look up, link to, and update experiment materials and ingredients without having to leave the ELN environment.

Going beyond the bench, having all data in one, cloud-based hub also makes it sharable and reviewable in real time so that productive collaboration can occur between access-approved chemists and formulators, product innovators and testers, process and production engineers, and decision makers.

SaaS data storage and management also saves on cost and resources and all software and data updates happen automatically so information is always fresh.

Accelerating informed decision-making with smart software

After the results are carefully documented, scientists are left with raw datasets. That data then needs to be digested and analyzed to extract meaning and guide go to market decision-making, particularly around key factors like performance and commercial scalability.

In some cases that decision might be straightforward (e.g., with uncomplicated materials like a polyvinyl acetate adhesives), but most of the time it is not that clear cut. Chemical production is becoming increasingly complex given shifting customer and market demands and mushrooming data volumes and observables alongside technological advancements in testing instrumentation.

Having innovative data management software that can analyze all results collected throughout R&D cycles helps identify things like structure performance relationships and optimizes the use of multi-parameter studies and scoring tools to assist in wisely choosing formulation candidates. 

Here’s how it works: selected files are imported into a data analysis system that uses mathematical modeling and statistics as well as customizable data processing workflows so chemists can begin to make sense of their multiparametric data. 

For example, a chemist designing a formulation for a new polymer-based surface coating for aerospace can identify key metrics like hardness, then upload their experimental results on scratch resistance and anti-deformation from tests on multiple formulations via the ELN. Performance is then illustrated for each formulation with scatter plots of hardness and density metrics simultaneously visualized and correlated. Data set outliers can be quickly identified and filtered out or kept in the analysis to re-consider at a later date.

The settings can then be adjusted to focus in on target performance characteristics (e.g., setting upper and lower boundaries for acceptable hardness and density measurements) so researchers can quickly identify formulations with values that fall in the acceptable range. They can then choose the most promising formulations and conduct a cost analysis bringing in commercial considerations to gauge feasibility. 

Additionally, with cloud-based data analysis software, approved R&D team members can view the analysis window and decide to make changes to the desired thresholds of the measured parameters in real time. This allows the scientist to pivot, generate, and review new results without having to wait for analysis data files to return to them.

By contrast, if chemists and formulation scientists lack digital tools, they have to sequentially analyze the hardness and density parameters in Excel and cost analysis would have to be done separately with colleagues and decision makers providing feedback via email or meetings—all costing time and resources. 

Additionally, data processing done with siloed, non-cloud-based software carries costs around reduced consistency, reproducibility, and untimely and less effective data sharing.

The future is now: Staying ahead of the curve with predictive analysis 

In addition to streamlining workflows and supporting data-driven decision-making with a SaaS approach, moving beyond the traditional design and test process to predictive analysis can help individuals and teams do wider exploration and spur stronger innovation. 

Accurately prediction of the properties of hypothetical formulations and molecules allows for optimum performance and outlier ideas to bubble up, which not only solves design problems at hand, but also perhaps identifies disruptive innovations to address unmet needs or suggest new ones. 

With chemical organizations increasingly adopting cloud-based data management, the industry is now poised to embrace digital transformation and help drive innovation to new heights. 

The capability of today’s next generation SaaS technology is up to the task, helping R&D teams take control of their data, their decision-making, and the evolution of their successful innovation cultures.