Sandvik uses data, analytics and AI to make mining more sustainable

“We see this digital solution as our way to provide visibility and sustainability to an operational environment that has long been hard to understand and forecast.”   

Esa Mattila, Reliability & Productivity Center Manager at Sandvik, is describing his company’s extraordinary monitoring solution – and how this is helping its customers to have greater control of their operations.   

“We are one of the leading providers of mining equipment in the world,” he says. “And since our customers operate in extremely challenging conditions, we know how critical it is that our machinery is always operating at its optimal efficiency.”

But this hasn’t always been possible for Sandvik and its customers.

“For many years, the mining companies we supply equipment to would only have high-level visibility on the state of their machines,” adds Mattila. “But this was never enough to really understand what was happening on their fleet at specific moments in time.”

Now, Sandvik has created a cloud-based service solution that uses data and AI to generate insights on the state of their machines to support the optimization of the operation of equipment. Implemented a few years ago, the technology is helping Sandvik and its customers to embrace digital transformation – bringing huge sustainability and efficiency benefits across their mines. 

“The solution we’ve created is all about using cloud technology to identify what is happening to the equipment we provide to our customers, as well as using the knowledge we have as an original equipment manufacturer to prevent issues from escalating and leading to system failures,” he says. “But there is so much more that you can do with AI and data. 

“You have infinite possibilities to make the industry much greener and more efficient than it’s ever been.”

From deep underground to the cloud

Sweden-based Sandvik Group is one of the leading mining and engineering companies in the world. Its Mining and Rock Solutions business area, where Esa Mattila works, is specialized in providing equipment, tools, parts, digital solutions and sustainability-driving technologies for the mining and construction industries.

“Our work at Sandvik focuses on manufacturing and delivering cutting-edge equipment that is used in rock processing and various mining operations,” he says. “These range from drilling, blasting, loading and hauling.”

Each of Sandvik’s machines is equipped with hundreds of model sensors that constantly collect values and parameters on the state of the equipment itself. “For years, Sandvik used to collect these parameters using local server solutions that, however, did not fully allow the company to leverage the power of Big Data,” says Mattila.

“We are committed to making the best of Big Data,” he continues, “so we recently decided to start storing all our machine telemetry data in a centralized location.  

“This centralized data storage repository soon became the foundation of our advanced analytics system.”

Bringing new light to the mines with the Remote Monitoring Service

Shifting to a Microsoft Azure environment was pivotal in helping Sandvik standardize and streamline its data. 

“As soon as we moved to the cloud, we started to get more and more data from our machines using IoT Hub and by moving to Azure Data Lake Storage and Azure Synapse Analytics, which in turn persuaded us to find additional digital tools to interpret it,” says Esa Mattila. 

“But in the end, we still felt like the full power of our data remained locked.

“That’s when we joined forces with Microsoft to create a new monitoring concept capable of converting data into actionable information.”

Called Remote Monitoring Service, the solution extracts information from the machines and uses AI to measure and analyze it. It is powered by a range of Azure components including Azure Synapse Analytics, Azure Databricks and Machine learning.

Sandvik uses Azure Databricks for data cleaning and preparation after moving it to Azure Synapse. With the data inside Azure Synapse, AI models and predictions are then created using machine learning. 

“We think it’s an incredibly powerful solution,” he says. “Not only because it helps us to keep track of our equipment, but also because it uses machine learning to understand our information and drive conclusions based on it.”

A gold mine for the entire sector

Sandvik is currently using its Remote Monitoring Service to drive insights in three crucial areas. These are predictive maintenance, operational efficiency and, most importantly, sustainability – driving substantial environmental gains by optimizing energy consumption and generating less waste. 

“Our solution can help us to identify when the machine is going to fail, understand its state and intervene before it actually fails,” says Mattila. “This allows us to deliver customers benefits in terms of productivity, safety, energy and fuel efficiency.”

"In turn, this helps us to significantly decrease operational costs, have fewer machine breakdowns and even reduce the amount of scrap metal and waste material we use.”

And there are huge operational advantages to the solution too. “When you have so much information on the state of your equipment, you can use it to improve operational efficiency. 

“This can lead to massive sustainability benefits: as our customers tell us, having a more efficient fuel burn allows them to achieve substantial improvements on machine utilization and reduce overall CO2 emissions.”

The ripple effects are significant not just for Sandvik’s operations, but for the entire mining sector. 

“As the process of electrifying machines continues, the mining industry is more and more in demand,” he adds. 

“This means that we can use our technology not just to improve our machines, but to go as far as building more sustainable mines altogether.” 

Bringing the initiative to the next level

Nearly a year since the start of the project, Sandvik is already seeing the positive impacts that its solution is having across its customer base. 

“We are currently measuring the success of our services based on the key performance indicators that our customers use,” says Mattila. “Many of them have told us they’re experiencing double-digit growth. 

“This is all thanks to the transparency and predictiveness that this service provides, which have allowed our operators to increase mine output in a more sustainable, safer and profitable way.”

The company has also started to use its data in the design phase of its machinery, as it attempts to address potential issues with equipment before it even reaches the mines. 

“Once you have the data, you can quickly integrate it with other systems and automate the entire business process,” he continues. 

“Using this combined approach of data-gathering and machine learning, we can use our insights to do virtually anything we want: from improving our machines to creating mining sites that are altogether more sustainable.” 

As they plan to integrate even more Microsoft solutions to the RMS, Sandvik and Mattila are hoping to make mining an increasingly data-led industry. 

“My motivation is to reach a point where our in-house systems and our customers’ systems become one, so that we can unlock the power of data even further,” he says. 

“Bringing even more efficiency and sustainability to our industry.”