The refractory industry has been around for centuries. During this time, the industry has been guaranteeing safety in high-energy intensive industries, primarily steel production, on a global scale. Without refractories, almost no process under extreme thermal conditions can function. The majority of the industry still adopts traditional practices, and the fact that it deals with raw material, takes place under extreme conditions, and lacks digital touchpoints means that it has not been at the forefront of digitization.
However, the market for refractories is expanding, and is expected to register a CAGR of 5% between 2016-2025. As demand continues to grow, there is mounting pressure to increase efficiencies and embrace technology in order to do so.
The industry’s traditional processes present a number of challenges in the age of globalization, including demand management, accurately analyzing production processes, stock monitoring, and supply chain management. As the mining sector and heavy industries form part of the Fourth Industrial Revolution, the refractory industry now has the opportunity to leverage data and technology to solve these challenges and drive innovation, but how exactly can it do this?
Due to the high-risk factor of the thermal processes where refractories function, product wear must be under constant surveillance. This is essential to guarantee the safety and quality of the product, and drive efficiencies within the overall process. Exchanging the product prematurely results in higher costs and unnecessary downtimes of the reactor. However, waiting too long to swap in a new product puts the safety of the workers involved at too high of a risk.
This is where in-situ wear analysis comes in: Computer vision technology can gather data to detect wear and mechanical stress in the refractory layer, which is able to “see” inside the reactor through optical or laser vision. This non-destructive approach provides relatively accurate results and allows companies to track exactly where repairs in the reactor have to be made before the next cycle or if a complete rebuild of the lining is necessary.
The pictures and other data gathered from in-situ analysis can be fed to machine learning (ML) models. Over time, the models will provide insights which allow companies to make accurate predictions on things like the running time of the product, and which parameters and conditions (such as temperature) inside the reactor are optimum to receive the desire product, be it steel, glass, or cement, while keeping the refractory layer at a minimum.
RHI Magnesita are pioneering efforts to use AI to learn from production data. Its Automated Process Optimization (APO) system receives all available data about the given production process, such as temperature changes, chemical processes, optical measurements, order cycles and planned maintenance work.
Using this information, along with empirical values and previous measurement results, AI-powered APO is able to make predictions about the maintenance and replacement of refractory materials. For example, the customer leveraging APO can plan maintenance work to take place between production peaks, thereby optimizing time and resources.
Network demand planning is one of the biggest challenges faced by the refractory industry. Much of the monitoring within warehouses and along the supply chain is still performed manually by humans, and information is stored in on-premise systems such as SAP, creating silos. The same goes for information from other parts of the value chain, including mining, raw material production, right up until replenishment at the customer’s site, and recycling and disposal of the product.
However, when integrated with an end-to-end cloud infrastructure, all of this data can drive newfound efficiencies for the industry. Data pulled from local warehouse monitoring combined with movement data, production cycle information and location tracking data can enable companies to anticipate where demand will arise and thus direct their product towards the right area at the right time. By adding a single identifier to every single unit or pallet via tracking methods (such as QR codes), companies can digitize their inventory data and monitor it as it moves through the supply chain.
Another challenge is integrating this data from all of the different production sites around the world. A centralized, cloud-based system is essential here to create data lakes where all of the relevant data can be securely stored, and remotely and securely accessed from anywhere, by any authorized person who needs it.
The refractory industry may not yet be at the forefront of Industry 4.0, but by generating new data with in-situ process analytics and utilizing existent data in cloud-based networks, it can close the information gaps within the supply chain, overcome legacy issues, modernize traditional processes, eliminate silos, and build smart, globalized networks that deliver crucial insights. With the help of an expert partner like intive, bringing these innovations into the fold and discovering new efficiencies is closer than you think.