Streamlining the chemical innovation lifecycle
Dr Michael Doyle, director of product marketing and principal scientist at Accelrys, looks at the role of IT in shrinking the gap between research lab and final product
Whether working to develop a flavour, a shampoo surfactant, a polymer coating or a flame retardant, companies that deal in speciality chemicals, formulations and materials face many challenges. Shrinking product life-cycles, global competition and pricing pressure from private labels are pushing chemical suppliers to get new innovations to market more quickly and cheaply than ever. Product complexity has reached an all-time high, just as the recession has prompted research organisations to slash budgets and resources.
Compounding these issues are increasing expectations related to quality, consistency, global specifications and environmental-friendliness, requiring more rigorous evaluation of the safety and stability of chemicals and formulations. Staying competitive in such a demanding environment means that companies need to start thinking about how to modernise their approach to R&D, starting with the tools and information technologies that support data integration, collaboration, analysis and process automation.
In 2011, the White House Office of Science & Technology announced the Materials Genome Initiative for Global Competitiveness, proposing 'a new national infrastructure for data sharing and analysis that will provide a greatly enhanced knowledgebase to scientists and engineers designing new materials.' The White House recognised, along with many in science and industry, that the time it takes to take to move an innovation from lab to market is too long, too costly and inefficient.
Fortunately, a new class of technologies that address the unique challenges inherent in R&D processes - such as lead identification, safety testing, specification management, production scale-up and more - are now coming of age. Here are several ways that companies can take advantage of information technology to streamline and speed the innovation lifecycle.
According to the Materials Genome Initiative white paper, extensive reliance on trial-and-error experimentation is a key contributor to the lengthy cycle times required to move a new product out of discovery, through development and on to production. Multiple rounds of experiments and testing can take months, even years, especially for complex and highly variable materials such as polymers, dyes, or coatings that that can be impacted even by small changes in chemistry, composition or processing steps.
Doyle - IT is crucial to shrinking the product development cycle
Additionally, these repeated cycles are poorly and inconsistently documented. To speed up discovery and development, organisations must be able to reduce the number of lab experiments needed, without hampering the ability of chemists and other research project participants to generate quality leads.
Technology that facilitates the modelling and simulation of chemical compounds and materials presents a compelling alternative to the use of trial-and-error experimentation alone. Already used widely in pharmaceutical research, software-enabled scientific modelling and analytical tools make it possible for researchers to design and test chemical compounds, mixtures and formulations in silico.
Instead of running numerous experiments in a lab, researchers can take advantage of these technologies to explore a broad range of ideas and better predict factors such as effectiveness, performance, stability, safety, toxicity, etc., before doing any chemical synthesis or live testing. This not only significantly reduces the time and expense that goes into lab experimentation and screening, it also supports improved innovation.
When product contributors can investigate more leads than would otherwise be possible, they are more likely to hit on a novel discovery. In addition, modelling and simulation can contribute to more sustainable R&D practices by saving resources that would have been used during extensive experimentation. It can also help organisations to design and test chemical mixtures that use greener source ingredients or catalysts cost-effectively.
Consider the experience of the R&D group at PPG, a diversified manufacturer of protective and decorative coatings, sealants, adhesives, glass products and industrial, speciality and fine chemicals. Using modelling and simulation software from Accelrys to identify candidate molecules capable of producing the desired properties, they cut product development time for a new industrial powder-coating and organic light-emitting diodes (OLEDs) in half, saving development resources and speeding time-to-market.
According to the case study description: "Each candidate that was conclusively shown not to work through in silico means can be seen as avoiding three to six months of experimental work. In addition to direct savings in avoided experimental costs, this approach raises the probability of finding an optimal end product because it was economical to search through and 'test' a much broader set of potential molecules for the OLED products."
Figure 1 - Demand-driven innovation
Several conditions need to be met for modelling and simulation technology to be useful to many stakeholders within the R&D enterprise. First, the technologies deployed must be simple enough to use so that even non-experts can take advantage of them. In the PPG case discussed above, the cross-disciplinary team working on industrial coatings that defined the properties required included marketing executives in addition to synthesis chemists and formulation chemists.
Secondly, the tools need to be capable of accessing and using multiple data sources, including historic experimental archives, current research and even publicly available information, to build the most accurate and complete models. Thanks to technological advances that facilitate collaborative data sharing and analysis, the power of these predictive tools can now be realised on a much broader scale.
Integrated information management
Data, in whatever form it is found, is the driving force behind every R&D effort. When critical information is not managed properly, project stakeholders cannot share it, valuable insights are missed and innovation processes quickly grind to a halt. The Materials Genome Initiative cited a lack of data transparency and integration across the product discovery-design-develop-manufacturing continuum as another key issue.
Part of the problem is that R&D data is extraordinarily complex and disjointed. Unlike the structured row and column-based data sets that are commonly processed through other types of enterprise information management systems, it includes unstructured information such as scientifically meaningful text, multi-scale models, images, outputs from sophisticated laboratory equipment and more.
This data exists in diverse array of formats, and is often 'siloed' within proprietary databases and IT systems. And the volume is enormous. Millions of molecular combinations may need to be explored to build a new alloy, polymer or catalyst, for instance, while tens of thousands of formulations may need to be screened in order to create a safe and stable hair dye.
Further complicating matters is the global span of the modern R&D enterprise, which reaches across scientific disciplines, geographic locations and organisational boundaries. Speciality chemicals manufacturers usually represent a single link in a value chain that also extends to the companies who design, develop and market the final product, such as electronic device manufacturers.
Both within and beyond the 'four walls' of an organisation, critical knowledge and processes need to move fast, so that suppliers can quickly adapt to the needs of their private label customers and the final product is more closely aligned to the chemical-level innovation that drives quality, sustainability and competitiveness.
What is required is an end-to-end, enterprise-level informatics platform to manage the innovation lifecycle. The platform should digitally facilitate the data integration, process automation and information sharing required for better compound management, for virtualised science like modelling and simulation and for richer collaboration. It should also function in a similar manner to solutions like PLM and ERP, but with the scientific depth needed to support the complexity of R&D metadata.
An enterprise-level informatics platform is needed to manage the innovation lifecycle
With the rise of cloud computing, service-oriented architecture and the use of web services and technologies that support advanced search and data mining, more streamlined 'e-innovation' is now a real possibility. Web services can, for example, be used to support 'plug and play' integration of multiple data types and formats without customised - and expensive - IT intervention. As data previously scattered throughout the R&D enterprise is made accessible across departments through a central informatics framework, it provides a 'single version of the truth' that drives a number of time, cost and efficiency benefits.
No matter where or how data was generated, numerous contributors to the product development value chain can use it, enhancing collaboration. Toxicologists can make their history of assay results available to formulators developing mixtures for a new cosmetic, for example, or chemists can work more closely with sourcing experts to ensure that the compounds they are developing in the lab are actually viable candidates for large-scale production.
In addition, information access is democratised, allowing 'non-experts' to take better advantage of highly sophisticated data, like models and simulations, which they may not have been able to use before without expert help. Processes that were previously fractured can be automated, speeding cycle times. Knowledge is more easily captured and archived, promoting re-use and reducing unnecessary rework.
In addition to accelerating innovation from the discovery of leads through production scale up, an enterprise informatics platform must finally be able to hand off critical data, in a useable, structured format, to PLM and ERP systems that govern product manufacturing, supply chain management and distribution. This 'last mile' completes the cycle from the research lab and final product.
Information technologies that allow users to take advantage of virtual science and that integrate data and processes across the end-to-end R&D enterprise can help chemicals companies address many of the challenges presented by a complex and demanding global market. By streamlining the innovation lifecycle, organisations will be much better prepared to compete and thrive.
From Online Issue: February 2012