Nothing but the Ground Truth

One year ago, in October 2023, our project partner NEO installed a Hyper Spectral Imaging (HSI) system in Leoben and provided us with training to generate a substantial dataset—a quite large dataset of over 150 GB—aimed at developing a classification model to support LIBS measurements for sorting our spent refractories.

This HSI system consists of a high resolution optical sensor operating in the 400-2500 nm range. It can be used to detect and differentiate materials based on their unique spectral signatures, which are influenced by mineralogy and bonding systems. The HSI sensor supports the LIBS unit in two primary ways:

1. Identifying Key Regions: It locates areas on the sample that are significant for classification, such as regions that are free from contamination.
2. Material Classification: It classifies spent refractories, thereby improving accuracy and validating classification decisions.

To achieve effective classification, the system must be trained using samples with known properties, referred to as ground truth data. This data serves as a benchmark for training multivariate and machine learning models. In the context of optical sensors, these samples are typically prepared under controlled conditions, ensuring that their spectral properties are well-documented and reliably reproducible. Ground truth data is crucial for training algorithms, as it provides a reference for the model to learn how to recognize patterns. However, as you might recall from previous blog posts, determining the exact material properties of spent refractories poses certain challenges. Fortunately, we recognized the importance of sampling early in the project and gained valuable insights into our feedstock, enabling us to provide well-characterized samples for our initial training.

For our sample preparation, we selected spent refractory samples from various customers and aggregates corresponding to specific sorting classes including defined common contaminants. We identified these classes either visually through well-known optical features or by chemically analyzing the bricks for confirmation. We then manually crushed the samples to ensure their integrity, ultimately providing images and data for about 2,000 pieces representing the sample shape to be placed on the conveyor belt in the sorting equipment.

Once prepared, these samples were analyzed using HSI. For data acquisition, each sample was illuminated, and the resulting spectral data was recorded as a reflectance spectrum. Consistency in data collection and meticulous documentation was crucial, so in total, we stored the aforementioned 150 GB of spectral data, along with an additional 4.5 GB of images documenting the sample origins, setup, and measurement settings.

After gathering the spectral data, we shared it with NEO to extract relevant features for classification. This process may involve statistical analyses or dimensionality reduction techniques, such as Principal Component Analysis (PCA), to pinpoint the most informative aspects of the spectral data. To aid in this, we also provided spectral data for our most common raw materials, helping to link spectral features to material classifications.

With the ground truth data and extracted features, we began the actual model training. By applying machine learning algorithms, such as Support Vector Machines (SVM), the model learns to associate specific spectral patterns with their corresponding material classifications. The final step involves validating and testing the model to ensure its effectiveness. This validation process requires a separate set of samples not included in the training phase, which assesses the model’s ability to generalize to new, unseen data—an essential measure of its practical applicability.

A year after the installation of the equipment and continuous exchange with NEO, we are pleased to report that we can classify certain sorting classes with an accuracy exceeding 95%. While fine-tuning remains necessary and the upscaling to actual equipment is still in progress, we are excited about the advancements made in our material classification capabilities. By focusing on chemical and mineralogical properties rather than visual features, we are not only minimizing subjectivity in our sorting process but we are also able to customize our sorting classes to further valorize our circular raw materials.

About CPI

CPI accelerates the development, scale-up, and commercialization of innovations in smart AgriFoodTech, energy storage, HealthTech, materials and pharma, while also catalyzing the adoption of advanced technologies and manufacturing solutions. These efforst aim to benefit people, communities, and the planet.  Through incredible innovation experts and infrastructure, experts at CPI look beyond the obvious to transform healthcare and drive towards a sustainable future.

As a trusted partner of industry, academia, government, entrepreneurs, and the investment community, CPI connects the dots within the innovation ecosystem to make great ideas and inventions a reality. Being part of the High Value Manufacturing Catapult, they facilitate access to world-class organisations to deliver transformation across industries and landscapes.

From the North of England and Scotland, CPI invests in people and disruptive technologies to invigorate economies, create circular supply chains and make our world a better place.

CPI Today

As of 2024, CPI has hit its 20-year milestone. Twenty years of innovation and impact, shaping the economy of the future with ground-breaking technologies.  It started with a clear strategy to leverage the UK’s knowledge and skills to ensure the economy remained competitive in the growing global marketplace.

Now with nearly 800 employees, CPI celebrates its growth as one of the leading influences on innovation in the UK. The organisation plays a crucial role in enabling the manufacturing sector, building national resilience, addressing health and well-being challenges, and supporting ambitious sustainability targets.

Through dedicated people and valued partners, the organisation has collaboratively transformed challenging scientific concepts into commercially viable products, unlocking almost £3bn in private investment.  CPI has a 60% SME customer base and is helping to transform industrial landscapes in the North East and beyond.

The best thing about innovation is that it never stands still, and CPI’s mission is not over. Innovation has been at the core of their work for the past twenty years and will remain the driving force for many years to come. If you are interested in knowing more, please visit www.uk-cpi.com.

The International Conference on Conveying and Handling of Particulate Solids (CHoPS) is widely regarded as one of the premier international forums for professionals in academia and industry to exchange ideas, share knowledge, and collaborate on innovations in the field of particulate solids handling and processing. CHoPS brings together experts who focus on the efficient handling of powders and particulate materials, fostering collaboration and sparking breakthroughs that advance industrial processes worldwide.

This year, as the 11^th edition of the CHoPS series, CHoPS 2024 took place at the Edinburgh International Conference Centre (EICC), from September 2^nd to 4^th, 2024, hosted by the University of Edinburgh in collaboration with an international organizing committee of industry leaders and experts. Over the course of three days, the event continued the success of previous years, offering an excellent platform for researchers and professionals to present innovative discoveries, discuss the latest industry challenges, and develop global partnerships.

The SINTEF research team, representing the ReSoURCE project produced an abstract, titled ‘Enhancing Recycling Efficiency of Residual Refractory Materials: Experimental Methods and Fine Particle Classification Techniques,’ which was selected for a full oral presentation. The presentation addressed the challenges of recycling residual refractory materials, particularly the finer particles (less than 5mm), which present unique difficulties for both environmental and industrial health due to their small size and potential to release dust.

Recycling efficiency is critical to the successful implementation of circular economy principles. However, the behavior of fine particles during recycling processes poses a significant barrier to fully utilizing reclaimed materials as secondary raw materials. This challenge is particularly pronounced in the refractory industry, where fine residues are often difficult to separate and reuse.

In their presentation, the SINTEF team emphasized the need to enhance fine particle classification methods to overcome this limitation. The talk focused on experimental approaches that highlighted innovative air-assisted classification techniques, including crossflow air classifiers and multi-chamber fluidized bed classifiers. These methods, which utilize differences in particle size and density, have shown promising results in enhancing material recovery efficiency.

The methodologies discussed in the presentation showcased the potential of direct sorting techniques to increase material recovery rates, significantly contributing to sustainable industrial practices. The audience responded enthusiastically to the team’s findings, recognizing their applicability not only to the recycling of refractory materials but also to other industries facing similar challenges with fine particle processing.

The presentation sparked a dynamic follow-up discussion on how the techniques developed within the ReSoURCE project could be adapted across multiple sectors, such as construction, mineral processing, and other industrial fields. This exchange of ideas underscored the value of multidisciplinary collaboration in developing solutions that can have broad applications, advancing the vision of a circular economy and sustainable resource management.

A  Global Leader in Refractory Solutions 

RHI Magnesita stands as a world-leading provider of refractory products and services, which are essential for industries operating at high temperatures, such as steelmaking, cement production, non-ferrous metals, and glass manufacturing. These industries require specialized materials that can withstand extreme temperatures, often exceeding 1,200°C, to protect equipment, enhance efficiency, and maintain safety in high-heat environments.

With headquarters in Vienna, Austria, RHI Magnesita has a broad global reach. The company serves as a crucial partner to industrial clients across five continents, offering not only high-quality refractory materials but also comprehensive solutions that span from mining to recycling. RHI Magnesita’s vertically integrated business model ensures control over the entire value chain, from the extraction of raw materials to the development and delivery of advanced refractory products.

A Powerful Merger

In 2017, RHI Magnesita was formed through the merger of RHI AG, an Austrian company with strong European roots, and Magnesita Refratários, a Brazilian company with a significant presence in Latin America. The merger combined the strengths of both companies, creating a global leader with unmatched expertise and resources. This strategic combination allowed the new entity to leverage its diversified geographic footprint, technological leadership, and comprehensive product portfolio to better serve a variety of industries worldwide.

The integration of the two companies also enhanced RHI Magnesita’s research and development (R&D) capabilities, positioning it as a leader in innovation within the refractory sector. The company invests heavily in R&D to continuously improve the performance and lifespan of its products, while also reducing their environmental impact. This commitment to innovation enables RHI Magnesita to remain at the forefront of the industry, setting new standards for high-temperature materials and processes.

Innovation and Sustainability

As a global industry leader, RHI Magnesita prioritizes sustainability in its operations and product development. The company is focused on reducing its environmental footprint through initiatives such as the development of refractory products containing circular raw materials. One of its key goals is to enhance the recycling of used refractory products, thereby minimizing waste and conserving resources.

RHI Magnesita’s sustainability strategy is also aligned with the broader global push towards reducing carbon emissions. The company is actively working to improve the energy efficiency of its production processes and reduce CO2 emissions across its operations. By promoting responsible resource management and developing innovative solutions, RHI Magnesita is well-positioned to support its customers in their own sustainability efforts.

Global Presence and Expertise

RHI Magnesita operates over 47 production sites and 8 recycling facilities globally, along with more than 70 sales offices, ensuring that it can serve customers efficiently wherever they are located. The company’s diverse geographic footprint allows it to meet the specific needs of regional markets while maintaining the highest international standards for quality and performance.

With a workforce of  20,000 employees, RHI Magnesita draws on its extensive global expertise to deliver solutions tailored to the unique challenges of each industry it serves. The company’s comprehensive service offering extends beyond products to include technical support, performance optimization, and long-term maintenance strategies for its customers.

RHI Magnesita is more than just a refractory supplier—it is a global partner dedicated to innovation, sustainability, and customer success. By leveraging its expansive global footprint, cutting-edge technology, and commitment to sustainability, RHI Magnesita continues to lead the way in high-temperature industrial solutions, ensuring safety, efficiency, and performance for industries around the world. For more information, please visit rhimagnesita.com and to keep up with the latest innovations and industry insights, join RHI Magnesita’s LinkedIn community

Having passed M24, it is a significant relief that the defined milestones have been achieved on time. Meanwhile, all work packages and many tasks within them have been initiated, leading to a tremendous acceleration in technical progress. As the technical coordinator, it is satisfying to witness such substantial work accomplished. However, it also means that I can no longer dive into every detail. Fortunately, I can rely on a great team, both external and internal, ensuring I receive the most relevant information to maintain a full focus on the bigger picture — bringing the world’s most advanced automated sorting equipment for spent refractories to life.

This milestone will soon be a reality with the completion of the construction and subsequent validation of Demo A, set to happen this year. While significant developments have occurred across various process units, the ultimate test will be the system’s machine performance, primarily reflected in its throughput when all units are integrated. It is crucial to identify the bottleneck in the system. Although we do not anticipate major issues with any unit, even minor discrepancies, like a unit operating at 92% instead of 95% efficiency, can have substantial economic impacts. Therefore, the next few months will be dedicated to several key steps, which we will update in our blog.

Here’s a simplified illustration of the steps we will focus on:

1. Performance of the Singularization Units: Ensuring all particles are well separated to be individually recognized by the system and handled without interference by robots or ejection mechanisms. With the singularization units ready, including the bunker, vibration feeder, and two conveyor belts, a parameter study will commence soon.

Figure 1. Exemplary top view on the conveyor belt after singularization, showing a satisfying result (left) and problematic grouping of particles (right).

2. Correct Classification of Particles: With the main sensor setups in place, we will closely examine particle classification. Initially, broad specifications for sorting classes will be defined, and then these will be refined to their limits

3. Gripping Performance of the Robots and Air Ejection Accuracy: Investigating the mechanical ejection process to ensure particles do not slip from the gripper or show a deviating particle trajectory, not landing correctly in their designated bins.

Figure 2. Picker robot gripping a particle. Images obtained at pilot trials at GeMMe Liège.

These trials will allow us to quantify throughput and calculate process economics, revealing areas needing further development and assessing the robustness of individual units. While there is a lot of work ahead, it is the kind of work that engineers and researchers genuinely enjoy. Stay tuned for our upcoming blog posts where we will delve into each of these steps in more detail.

In environmental science, leaching tests play a crucial role in assessing the potential impact of materials on the environment. But what exactly is a leaching test, how does it work and why did we use in project ReSoURCE?

What is a leaching test?

A leaching test is a laboratory procedure used to evaluate how substances, such as metals or chemicals, can be released from a solid material (spent refractories in our case) when it comes into contact with a liquid, typically water. This process mimics natural conditions where rainwater or groundwater might interact with waste materials, soil, or industrial by-products.

How does it work?

The basic principle involves placing a sample of the material in a container and then adding a specific amount of liquid. The mixture is then agitated for a set period to simulate natural leaching conditions. Afterward, the liquid, now called the leachate, is analysed to determine which substances have been released from the solid material and in what concentrations.

Why did we perform leaching tests in project ReSoURCE?

Despite the above-mentioned reason (potential impact on the environment) we also wanted to analyse the behaviour of so-called impurities. When we talk about impurities, we mostly mean alkaline elements like K₂O, Na₂O, Cl^–, and SO₃^– which hinder further recycling by negatively influencing refractory properties. To remove these elements’ further treatment (e.g., washing) is necessary.

Leaching tests are often combined with aqua regia digestions, which provide information about the chemical content. This analysis comes with some limitations as some materials do not dissolve completely in aqua regia (e.g., some silicates). Nevertheless, by comparing the results of these two methods we can see how good a treatment method like washing could perform.

Leaching tests have a lot of variation like for example pH-depending leaching tests, where we analyse the leaching behaviour at different predefined pH.

Applications of Leaching Test Results

The results of leaching tests are vital for several reasons:

• Environmental Safety: They help predict the potential for contaminants to migrate into groundwater or surface water, posing risks to ecosystems and human health.
• Regulatory Compliance: Industries use these tests to ensure that their waste management practices comply with environmental regulations.
• Remediation Planning: Leaching tests guide the development of effective strategies for cleaning up contaminated sites.

By understanding the leaching behaviour of materials, scientists and policymakers can make informed decisions to protect the environment and public health.

The image below shows an overhead shaker loaded with five samples, set to undergo a 24-hour shaking process.

 Sample image. Photo credits: Bettina Ratz.

SINTEF is one of Europe’s largest research institutes, with multidisciplinary expertise within technology, natural and social sciences. As a Norwegian R&D partner, SINTEF strives to contribute to value creation and increased competitiveness within the public and private sectors. With around 2000 employees, SINTEF delivers applied research, innovation, technological development, knowledge, and solutions for customers across the world.

In the ReSoURCE project, SINTEF is participating through its Industrial Process Design (IPD) group located in Porsgrunn. The group has been working with powder and particulate technology for more than 40 years and successfully carried out several R&D projects funded by EU, the Norwegian Research Council and various industries. It has been serving several industrial sectors such as oil & gas, metals, minerals, engineering, chemical, fertilizers, FMCGs, pharmaceuticals, food, and feed. SINTEF IPD provides R&D services for the entire value chain of the processes, i.e., from lab experiments to industrial scale implementations via pilot tests.

SINTEF IPD aims to foster sustainable process industries by integrating state-of-the art process technologies with advanced digital solutions. Primary goals of its initiatives include enhanced safety and improved resource, energy, and cost efficiency across various industrial processes, particularly those involving powders and particulate materials.

The powder handling and processing operations served by SINTEF IPD activities are presented in Figure 1.

Figure 1: Powder handling and processing activities at SINTEF IPD.

SINTEF IPD hosts the following pilot and bench scale test facilities:

• Pneumatic conveying pilot test rigs (dilute & dense phase)
• Powder Fluidisation test rigs
• Powder mixing units (batch & continuous)
• Sand blast type erosion tester
• Particle Comminution-milling
• Equipment for air classification & powder sorting

Additionally, its infrastructure to characterize various mechanical and materials properties of powders is as follows:

• Particle Properties (size distribution, densities, surface areas, flow properties)
• Flow properties (Jenike shear, uni-axial, angle of repose)
• Particle Shape Analyzer
• Revolution Powder Analyzer
• Evolution Powder Tester
• Sieving machine (120 µm – 2040 µm)

Some pictures of SINTEF IPD’s experimental facilities are presented in Figure 2.

Figure 2: Experimental facilities for powder processing and characterization at SINTEF IPD.

In recent years, the group has been involved with the following long-term projects:

• Refractory Sorting Using Revolutionising Classification Equipment (HEU Project GA No. 101058310)
• Open Innovation Digital Platform and Fablabs for Collaborative Design and Production of personalised/customised FMCG (EU H2020 Project, GA No. 870148)
• Improving Efficiency of Offshore Drill-cuttings Handling (NFR, PETROMAKS Project No. 234162)
• Effective Handling of Bulk Solids with Focus on Reduction of Erosion and Scale Formation (NFR BIA Project No. 247789)

In the future, SINTEF IPD intends to accelerate its activities on developing solutions for circularity of resources in process industries.

As a proud partner within the ReSOURCE project, we at CPI catalyse advanced technologies and manufacturing solutions that benefit people, places, and our planet.

Our involvement in the ReSOURCE project is to develop alternative, non-refractory, uses for the leftover spent refractory materials generated during the breakout of the refractory linings. There are a variety of potential alternative uses for these spent materials for value-added purposes. At CPI, we have been focusing on using the spent material as additives in polymer composite systems. We’ve also developed cost-effective methods to mill the materials to a suitable size for incorporation in polymer systems.

A composite material is a combination of two materials with different physical and chemical properties. When they are combined, they create a material that is specialised to do a certain job, for instance, to become stronger, lighter, or resistant to electricity. They can also improve strength and stiffness. Common fillers and reinforcing agents in composites include calcite, kaolin, carbon black, glass, and carbon fibres. The leftover refractory brick residues, because of their service life history, are expected to impart good thermal resistance and enhance mechanical properties when used as a filler at an optimum dosage in the formulation.

The detailed investigation of the material properties, aimed at the specific usage in combination with a polymer matrix, has been articulated in technical information literature. Along with the technical information literature, a safety data sheet for the end user has also been developed as part of the material passport. Any end user currently utilising marble powder and/or cheaper clay material as a filler may benefit by replacing either part or all, of the filler material.

A careful study of the material passport will generate quite a few good ideas about the specific end-use properties that can be expected from the composite material.

Explore the technical data sheets now available in the Knowledge Vault section under ReSoURCE results for detailed insights.

As presented in the blog earlier this summer, the assembly and commissioning of the demonstrator sorting system for refractory bricks has already started at LSA’s facilities in Aachen. As part of the ReSoURCE project, another demonstrator, handling refractory material with less than 5 mm particle size ,will also be designed and tested. This demonstrator is smaller than the one sorting bigger bricks and will have a more scientific approach. The demonstrator will be used to learn how sorting and handling of the fine fraction of spent refractories can be optimized.

The design phase of an industrial scale demonstrator has several stages, and many partners from the consortium are contributing with their knowledge and experience. At SINTEF, we are specialized in powder and particulate material handling and technology. Our contribution to the design of the demonstrator has been to recommend separation/classification technologies and to do calculations of design parameters for those systems.

Experimental testing is a useful method to test specific separation methods and to evaluate how efficient the method works before deciding on the final design. This is especially useful for complex materials where physical and chemical data is unknown and where a simulation exercise would be too complex. Initial powder characteristics analysis in the laboratory is the first step to get to know how the powder can be handled at industrial scale. The particle size distribution, flow properties and density are parameters used to decide which handling and classification equipment that could be suitable for the specific material.

Based on experience from powders with similar characteristics, available information in the literature and available test equipment, some classification methods were chosen for further testing. Real spent refractory material has been provided by RHI Magnesita and a representative test setup was then prepared. Different experimental parameters such as throughput, air velocities and feeding rates were tested, and samples from each test was used to evaluate the separation efficiency both for chemical differences and particle size separation.

When the parameters for the throughput and separation process was evaluated, design parameters for rest of the demonstrator could be calculated. That included feeding systems, transport systems and collection system for the different powder streams. Using this input, everything can be put together in drawings and the search for suitable and available parts and suppliers can start.

Figure1: Draft of the most important elements and process steps in the demonstrator for refractory fines (<5 mm).

Crowdhelix is an open innovation platform that forges links between an international network of excellent researchers and innovative companies, enabling them to plan, deliver, and exploit pioneering collaborative projects and value chains. As Europe’s leading impact acceleration partner, Crowdhelix offers a unique networking, clustering and value chain creation platform managed by Impact management experts from across Europe and beyond.

The platform employs an interactive approach called the “Helix Impact Model,” which facilitates the dissemination, exploitation, and communication within EU-funded research projects. Each Helix is designed to bring together experts, RTOs, SMEs, policy makers, and other relevant actors for collaboration and impact in research and innovation. Helix communities enhance collaboration among different stakeholders along the value chain of the most pressing topics and technologies, allowing adaptation to various scientific and innovative fields, and boosting different ecosystems in the open innovation landscape.

The Helixes are a crucial component of Crowdhelix’s three-step Impact Acceleration Approach, which includes stakeholder analysis, matchmaking and clustering, profiling and connecting innovation to global end-users. This is achieved by using a machine learning driven ‘recommender engine,’ to match collaborative partnerships. Crowdhelix currently hosts 54 active Helixes where users can profile themselves, their teams, their organisations, and post opportunities to collaborate. An intelligent recommender system then matches these opportunities with the most suitable prospective collaborators, using natural language processing and machine-learning.

Any researcher, student or expert of an organisation that is already a member of the Crowdhelix Network and has a valid email domain provided by their organisation can register, post about collaborating, or connect with other organisations searching for expertise in different areas of research. Users can proactively post and respond to collaboration opportunities within the Helixes, and target specific Horizon Europe funding calls drawn directly from the EU’s database. In that sense, Crowdhelix makes it easy for diverse businesses, universities, and research organisations to find common ground, and forge strong partnerships so they can develop competitive project concepts and proposals targeting Horizon Europe funding.

The platform’s emphasis on impact and sustainability has led to the development of groundbreaking technologies and solutions across various sectors. Crowdhelix has facilitated numerous successful projects, bringing together top-tier researchers and industry leaders to tackle complex challenges.

Being a global collaboration platform for researchers and innovators, Crowdhelix’s members have collectively been awarded over €11.79 billion of funding under the EU Horizon programmes, to deliver 22,246+ pioneering projects. Learn more about successes that Crowdhelix has been able to facilitate on their outcomes page as linked here.

In addition, Crowdhelix delivers over 40 sector leading live and virtual events per year. Attending Crowdhelix events allows researchers, experts and students to focus on research and innovation domains and discover new opportunities and partnerships, as well as enabling connections with other expert members of Crowdhelix.  Looking to the future, Crowdhelix plans to broaden its network and improve its services to support even more ambitious and impactful projects. The platform is set to be a pivotal player in the continuous evolution of the European research and innovation ecosystem, driving progress and delivering value to all stakeholders involved.