This three-part series starts from a basic insight: through advances in digital manufacturing, raw materials are fast becoming intelligent assets. Thought of another way, material flows are becoming information flows. In this series we explore the implications for the circular economy. In part one we investigated the technological advances that are encoding intelligence into materials. Here we look at the trends in storing, communicating, and using materials data. Part three will turn to the impact on material supply chains and business ecosystems that result, and discuss the business models that stand to benefit from emerging trends.
In Part 1 we explored how raw materials are becoming intelligent assets – how we are encoding information into a material’s composition and structural form. In this article we examine the tools that allow the use of these new types of data, asking the important questions of who has access to materials data, who produces it, and who consumes it. Answering these questions will help us use advances in materials science to accelerate the transition to a circular economy.
When data is organised, it becomes information, but it is only when we understand how to use information that it becomes knowledge.1Machlup, F. (1983). Semantic Quirks in Studies of Information, in Machlup, F. and U. Mansfield (eds.) The Study of Information: Interdisciplinary Messages (p. 641). New York: John Wiley & Sons, Inc.
Data may be neutral, but the ways it is organised into information has implications for who can access and use it. Some of the most well documented and intractable problems facing the circular economy transition involve a lack of access to useful information about materials. Many initiatives are focused at making information on materials more transparent, for example throughout product value chains. Those with the tools to turn data on materials into useful information and new knowledge may hold the keys to a circular economy.
Let’s look first at the trend in how data on materials is stored, communicated, and used in manufacturing. Data on materials is stored in large, searchable databases that usually contain information on the chemical composition of a material, and how it behaves under certain environmental conditions. Many companies license access to material databases. An example is Granta, one of the market leaders that offers a massive repository of data on materials for manufacturing, coupled with sophisticated analysis software, to mainly industrial users. Another large database is MatWeb, which offers free materials data targeted at engineers and industrial users. There are many other material property databases available, with some tailored more to specific industries or academic uses.
Database companies work hard to make sure they have the latest materials in their system, and in doing so secure their position in the marketplace.2For example, Granta and MatWeb are both integrating 3D printing materials into their databases: Granta’s recent partnership with Senvol (www.senvol.com) has resulted in a growing list of 3D printing (i.e. additive manufacturing) materials being integrated into its set of tools and services.
However, their market dominance is being challenged by the very nature of the new wave of materials data. This data is far more complex and multilayered than traditional engineering data that has powered manufacturing for the last century. For 21st century manufacturing, engineers increasingly need to compute the properties of a material at nano and micro millimeter scales, and model how these properties interact with the object’s structural form to produce the characteristics of the final product.
The volume and complexity of such data is transforming material science into an information science. This is where the tools of ‘big data’ come in, with smart search algorithms and advanced analytics that help make sense of the messy heap of available data. If big data is the new black gold in today’s economy, data on materials is no exception. One example of a company on the cutting edge of this trend is Citrine Informatics, a data analytics platform for materials development, selection and sales. Citrine uses smart algorithms to mine academic literature, public data, and its own private sources to help companies identify materials suited to their design needs and challenges. Founded only in 2013, its data-driven approach signals a new movement in understanding materials as intelligent assets.
Materials as a big data science is also finding support in the corridors of power. In the United States, the Material Genome Initiative spans agencies including NASA, the US Department of Defense, and the National Science Foundation, with $500 million invested. The initiative is charting a path towards materials innovation that includes the Materials Project, a platform for computational material science that draws from a database of all known material properties. With this wealth of information, researchers can data-mine scientific trends in material properties and identify promising new materials to experiment with in the lab, which could help overcome some of the circular economy’s tricky design and materials challenges.The Materials Genome Initiative also includes the Materials Resource Registry, an initiative that aims to unify the world of fragmented scientific data on materials and make it easily searchable and accessible. The registry is built using open source frameworks and is open access, another difference from traditional licensing models that keep data within relatively closed industrial communities.
It remains to be seen how new big data tools for materials development and manufacturing are impacting the circular economy. As researchers and companies speed up the development of materials, we may see a massive expansion in the number of material compounds in use. This may cause considerable problems for cycling these materials. However, such tools may also be adapted to identify materials at the end of their lifecycle, creating a big data revolution in recycling and waste management industries. Without investment in such tools, the infrastructure needed to process and cycle materials back into value streams after the end of their useful life may lack capacity to keep up with the production of materials fuelled by big data science.
Following on from these big data initiatives is the second theme – how data on materials is communicated and shared. The Materials Resource Registry relies on a Materials Data Curation System that allows data to be captured and shared, and fortunately the system is flexible enough to include attributes of the material relevant to circular economy actors, like recyclability. Given that these data systems are multi-sectoral in scope and well financed, it will become increasingly important to integrate them into circular economy initiatives.
There are also new developments underway to standardise the communication of material properties at the micro and nano scales. With the help of 3D printing, we are now able to layer materials in patterns at the micron scale, creating material ‘microstructures’. Just like other forms of information, it is becoming important to create a common format for organising and communicating this data. Researchers at the German Technical University of Aachen have tackled this challenge by creating a common format for describing three-dimensional microstructures of materials akin to the jpeg format for exchanging digital pictures. Interestingly, these researchers aim to extend their data system to capture changes in the material over time. This could allow the tracking of detailed material property information over a product life cycle, enabling greater knowledge over the potential ways to cycle these materials back into productive use.
Storing and sharing data on materials is only useful if we can make use of the data to make things. In particular, we must uncover how best to use these new data tools to make things for a circular economy. Today, ‘making things’ with material data takes place in digital and physical worlds, and the two are becoming fundamentally entwined through the design process, with materials database becoming increasingly linked with digital design and manufacturing platforms3For example, the Granta materials database can be linked up to Autodesk Inventor (http://www.autodesk.com/products/inventor/overview), mechanical design and 3D CAD (Computer Aided Design) software. MatWeb also offers exportable datasheets on material properties to CAD/CAM (Computer Aided Design/Computer Aided Manufacturing) programs like Autodesk (http://www.autodesk.com/solutions/finite-element-analysis) and Solidworks (http://www.solidworks.com/sw/products/simulation/finite-element-analysis.htm), two leaders in product design, simulation and manufacturing software.
. This means that data on material composition and behaviour can be integrated into the workflow of product design. Using Finite Element Analysis (FEA) designers can simulate how their product behaves under certain stresses and strains expected in the real world before it is physically manufactured4FEA software computes where the stresses and strains are concentrated, giving useful information to a designer who may need to strengthen parts of the design to make it more robust. Other forces like mechanical vibration, fatigue, and fluid flow can also be modelled.
. For example, you can model how a digital 3D model of a chair behaves in response to the simulated strain of a person sitting on it.
FEA has been a cornerstone of engineering for decades, but the advent of 3D printing has radically opened up the potential to design objects that are tailor made to withstand particular stresses and strains encountered during their use. An example is the Airbus hinge bracket, 3D printed using Direct Metal Laser Sintering (DMLS). Compared to the original part design, the 3D printed part has an optimal geometry for its mechanical function, and the more economical material design can reduce the weight of a plane by 10 kilograms. These software platforms are increasingly becoming widely accessible beyond industry and engineering communities. Autodesk offers many of their products free to students, and both Solidworks and Autodesk have released free professional versions to over 1000 Fab Labs worldwide, opening the doors for makers everywhere to do sophisticated materials design. High powered tools to use material data for making things is increasingly at our fingertips.
At present, the uses of these tools rarely go beyond material efficiency. This is changing, with the new wave of ‘generative software’ that challenges the very fundamentals of product design. Generative software harnesses the power of big data while taking its cue from nature. It allows the user to set high-level design constraints, such as material type, cost, and specific functions, and then ‘aids’ the user by running intelligent algorithms to generate a large diversity of possible product solutions that meet the pre-specified design constraints. This process mimics the process of evolution, where organic forms evolve to serve specific functions, all within the ‘design constraints’ of an ecological niche. This makes generative software a powerful ally of biomimetic design.
Generative software may become a critical tool for designing high performance products within circular economy design constraints. New products may be designed from a simple palette of recyclable materials, using the software to evolve complex structural forms (like internal lattice structures and topologies) to meet particular performance goals. Autodesk is the leader in this new paradigm with their Project Dreamcatcher. Already, Autodesk Within and Nastran software, allowing the iterative design of internal lattices and surface topologies, signpost what is to come.
Considering these trends, it is clear we are facing a world awash in materials data. We must respond to the changing landscape of materials data with initiatives that take full advantage of the tools of data science. Many advances are being made in how to organise and make use of it, and access to this data is opening up. This is good news for the circular economy. Industrial producers, designers, and makers everywhere may have the tools to evolve designs with smart materials that outperform today’s chemically intensive recipes. Information sharing protocols and sophisticated databases may make information loss over product lifecycles a thing of the past.
Much of this promise rests upon new business models and integration of circular economy aims into material data tools. We need to investigate the opportunities and dangers that materials as a big data science provides for circular economy initiatives. If we find the right points of leverage, raw intelligence may rapidly accelerate the transition to a circular economy.
References [ + ]
1.
↑
Machlup, F. (1983). Semantic Quirks in Studies of Information, in Machlup, F. and U. Mansfield (eds.) The Study of Information: Interdisciplinary Messages (p. 641). New York: John Wiley & Sons, Inc.
2.
↑
For example, Granta and MatWeb are both integrating 3D printing materials into their databases: Granta’s recent partnership with Senvol (www.senvol.com) has resulted in a growing list of 3D printing (i.e. additive manufacturing) materials being integrated into its set of tools and services.
3.
↑
For example, the Granta materials database can be linked up to Autodesk Inventor (http://www.autodesk.com/products/inventor/overview), mechanical design and 3D CAD (Computer Aided Design) software. MatWeb also offers exportable datasheets on material properties to CAD/CAM (Computer Aided Design/Computer Aided Manufacturing) programs like Autodesk (http://www.autodesk.com/solutions/finite-element-analysis) and Solidworks (http://www.solidworks.com/sw/products/simulation/finite-element-analysis.htm), two leaders in product design, simulation and manufacturing software.
4.
↑
FEA software computes where the stresses and strains are concentrated, giving useful information to a designer who may need to strengthen parts of the design to make it more robust. Other forces like mechanical vibration, fatigue, and fluid flow can also be modelled.
The post Raw intelligence: how data flows work, and why they matter appeared first on Circulate.