Kennametal pushes manufacturing into the future with IoT and Microsoft cloud solutions

At the Kennametal factory in Orwell, Ohio, Plant Manager Frank Constantine looks out over the shop floor where skilled operators manage a fleet of machines that produce metallic cutting, milling, and lathing parts for industrial applications. “At any given time, we have millions of parts at some stage of production in this plant,” he says. “In the past, we tracked them by logging their progress on paper forms. As you can imagine, it could be very difficult to manually locate an individual production job among all that inventory when our data was stuck in a pile of spreadsheets.”

From order tracking to machine diagnostics, the Orwell plant used to have a lot of manual processes that limited production capacity. But these days the plant produces at a whole new level, thanks to Kennametal’s investment in the Microsoft Azure cloud platform, Microsoft Power BI, and Internet of Things (IoT) technologies like Azure IoT Hub.

“Not only do we have a lot more data about our production processes, we have the tools we need to put that data to good use,” says Constantine. “When we need to find an individual order, we have that information at our fingertips. And now that we know when one of our presses isn’t running the proper tonnage or isn’t manufacturing at the proper speed, we can analyze those issues and address them in real time to achieve all sorts of operational and other improvements. With Power BI, we make decisions based on data, rather than gut instinct, and we have insights into our work that just weren’t possible before.”

The importance of information and cloud capabilities

The parts made in Orwell and at Kennametal’s other factories are used across a wide range of industries, from aerospace and transportation to energy and construction. The company serves customers in more than 60 countries, and its products and solutions help them to run longer, cut faster, and machine with greater precision. Kennametal has a corporate vision based on ongoing innovation and continuous transformation, and the company aims to set the highest possible performance standard for its industry.

To achieve that goal, Kennametal has embarked on a multiple-year Simplification/Modernization effort to turn its manufacturing plants into the factories of the future. The changes at the Orwell plant are one of the first steps on that journey—a venture driven, in large part, by data.

“Data is becoming a critical asset for every organization,” explains Brad Zorn, Director – Digital Manufacturing and Enterprise Analytics at Kennametal. “How we gather data and make sense of it determines our ability to make information-driven decisions that accelerate business growth and let us respond quickly to customer needs.”

As Kennametal embarked on its modernization initiative, the company quickly realized that its existing on-premises infrastructure could not accommodate the vast quantities of data produced by a modern factory or provide the analytical resources necessary to derive business insights from it. After evaluating options from several cloud providers, Kennametal chose to migrate its IT systems to Azure.

“With Azure, we get the instant scalability of the cloud, along with high availability and great performance,” says Dan Berlin, Global IT Director at Kennametal. “We also get access to the services and solutions that we need to put our data to work for us, from analytics tools to sophisticated AI and machine learning. We see Azure as a key enabler for our success.”

Detailed data drives rapid responses

Back on the factory floor in Orwell, Constantine looks up at a large monitor with a dashboard displaying the current operating status of the manufacturing machines. Kennametal uses Power BI to unify its data and give employees up-to-the-minute analytics. Red and green indicators show Constantine which individual machines are within operating specifications and which may have a problem.

“Before the modernization started, someone needed to walk around to every machine and check it manually, which isn’t as simple as it sounds,” he says. “The machines normally stop working repeatedly throughout the day for cleaning and setup, so our status checks involve more than just looking at whether a machine is running—we need to know exactly what it’s doing and for how long.”

The machines at the Orwell factory now contain IoT sensors that collect operational and production data to help employees better understand how well those machines are working. The machines are connected to Kepware servers that transmit the data through Azure IoT Edge to IoT Hub. Kennametal also uses Azure SQL Database with the single database deployment option for short-term storage, Azure Stream Analytics to help process the IoT information, and Azure Data Lake to store data for the long term and extract it for dashboards and reports.

“With IoT Hub, we have a scalable platform for ingesting all the incoming information. And with Stream Analytics, we have a scalable platform to move the data where it needs to go and transform it on the way,” explains Zorn. “We use SQL Database to quickly spin up resources without the need to order hardware for a particular site. In the past, it would have taken months of purchasing and deploying hardware to make that happen. Now we just pay for the computing power we use, and it’s available whenever we need it.”

This on-demand computing power makes it possible to quickly understand manufacturing exceptions. If a machine spends too much time cleaning, generates too much scrap, or produces substandard parts, it’s now much easier to trace the problem to its cause and fix it. Constantine recalls one instance in Orwell involving an excess scrap issue with a batch of parts produced on a machine that uses carbide powder as a raw material. 

“One of our supervisors detected the problem on a Power BI dashboard, and we used data analytics to trace that particular batch of powder back to the plant it came from,” he says. “From there, we determined that the quality issue was due to a single operator running a process out of specification for a brief period on a single day. Before the modernization, it could have taken us weeks of searching through spreadsheets and databases to determine something like that. But now we get answers like this in a couple hours.”

A network of connections for a company and its customers

Having better data in hand helps Kennametal advance its modernization goals and its plans for connected smart factories. “In addition to data, the factory of the future will be focused on people,” says Tom McKee, Chief Information Officer at Kennametal. “It’s about understanding what information our employees need and getting it to them when they need it so they can make timely decisions.”

Kennametal accomplishes this by using Azure and Power BI to connect machines and people through mobile devices or whichever medium employees need to have crucial information at their fingertips. From there, the network of connections grows outward.

“We started with a single smart factory, and from there we’re building out a network of smart factories that are all connected,” says McKee. “Ultimately, we aim to connect this network of factories to our customers. That gives us an intelligent enterprise throughout the entire end-to-end value chain, connecting the supply side—the factories—with the demand side of our customer needs.”

Adds Mike Shields, Executive Vice President, Founding Partner and Industry Strategist at eLogic, an important Kennametal technology partner involved with the factory modernization effort, “By introducing new technologies and solutions, Kennametal can gain actionable insights that very likely would never have come to light before. Together, we’re making correlations in data and surfacing new opportunities for the business. By bringing in Azure Machine Learning, there are even more exciting possibilities for predictive analytics and proactive decision making related to machine maintenance and product quality.”

Machine Learning is able to generate insights into otherwise hidden material science improvement opportunities. “One example is powder quality within Press Operations, which are a key production stage and determinant of throughput and quality,” says Shields. “Similarly, by capturing the shop floor Lead Board data in Power Apps and ingesting it into Machine Learning, we can identify cross-cell and cross-plant improvements that further optimize overall quality and performance.”

Kennametal has found great value in Power BI as a tool for making sense of its data within its first connected factory. The company is deploying additional solutions like Microsoft Power Apps so that it can operationalize data through applications to realize even more value. “For example, we’ve taken the data we’re collecting and used Power Apps to quickly create an app that alerts operators if something goes wrong,” says Shields. “Now operators can move freely throughout the plant to take care of tasks, with the confidence that they’ll receive a notification right away if a machine has a problem. And the alert automatically escalates if they don’t address it within a certain amount of time. Other apps are in the works as well—using Power Apps is a great way to quickly get solutions into employees’ hands.”

Business benefits and everyday enhancements

Kennametal’s emphasis on collecting and understanding data is paying valuable dividends at the Orwell plant. In one case, plant management used the data captured in Power BI to identify the need and successfully advocate for a capital expenditure project. “Having a better handle on data through Power BI makes it easy for us to measure a problem, document it, and justify the solution,” says Constantine.

The data and IoT effort has also made it possible to ramp up production at Orwell. Kennametal has improved machine setup times at the plant by almost 50 percent and also increased press availability by 50 percent. As a result, the plant now produces significantly more parts in certain areas than before with the same staff. And all these process improvements make life better for employees.

“I’m frankly excited about the opportunity to see my entire plant in Power BI at the click of a button,” says Constantine. “Traditionally, I’ve had to walk around the whole facility to get that information. Now I’ve got one screen where I can see how each department performs and what metrics they’re attaining. I can direct my attention to where it’s most needed.”

In these modernized plants, Kennametal sees important benefits for employee empowerment and engagement. “In line with a principle we call ‘ideas in motion,’ we want to use technology to encourage employees to submit ideas for enhancements and improvements,” says Zorn. “We also expect that having Power BI dashboards for machine status and other indicators will make it easier to cross-train operators on new equipment and give them more responsibility in a way that sets them up for success.”

For Tamala Ellsworth, a PVD Setup Operator at Kennametal, the changes have added an extra spark to her day. “I feel like I get a lot more done now,” she says. “And because it’s easier to see on the Power BI dashboard exactly what I’m producing, it inspires me to want to produce even more. There’s kind of a friendly competition between shifts to see who can beat whom. I’m looking forward to learning more about the technology so I can do an even better job.”

Find out more about Kennametal on Twitter, Facebook, YouTube, and LinkedIn.