KANSAS CITY, MISSOURI, US — Scheduled maintenance keeps industrial systems up and running and provides a good baseline, but it may not catch every issue and isn’t always the most cost-effective, said Jim Neawedde, food and beverage segment manager for ABB, Inc.

During his presentation on predictive maintenance at the 2023 Grain Elevator and Processing Society’s (GEAPS) Exchange in late February in Kansas City, Missouri, US, Neawedde asked: “What if it was possible to detect issues before they cause downtime?”

Predictive maintenance makes it possible to evaluate what’s happening in real time to address issues before a system goes down. It can help assess the health of individual components, detect issues with overheating, vibration or energy consumption. 

Predictive maintenance can help address some of the key challenges in the industry such as costs, safety of the product and in the supply chain, product quality and production reliability and sustainability. 

“If you’re not producing a quality product, you get a bad rep in the industry, which is not a good thing,” Neawedde said. “We want to have our systems up and running; we want to leave a legacy.”

Maintenance histories

There are multiple maintenance strategies, starting with corrective maintenance or run-to-failure. With this strategy, maintenance is only done when problems occur. The probability of failure is not under control, and it has the highest risk of production and service loss.

Time-based monitoring uses a schedule for when maintenance is performed. The probability of failure is under control just after the recurrent inspections. It also has high maintenance costs due to the recurring equipment inspections, Neawedde said.

Preventive maintenance is regularly scheduled, either using time intervals or usage (operations/cycle count) as a trigger. It can be enhanced with root-cause analysis and troubleshooting instructions. Condition-based maintenance is preventive maintenance supported by condition monitoring of the asset, with basic diagnosis on read values. There are less maintenance costs due to the reduced number of inspections. 

Predictive maintenance combines various sensor readings (condition monitoring), sometimes external data sources, and performs powerful analytics on thousands of logged events/data. It provides a continuous prediction of probability of failure and remaining life. It also has the lowest maintenance costs.

“You can see because of the way I’ve been abusing this piece of equipment, the expected lifespan is going to be shortened, or I’ve been taking really good care of it, I’m going to get longer life out of my equipment,” Neawedde said.

Digital technologies

Transitioning to predictive maintenance will require adoption of digital technologies. The agriculture industry is on the lagging end of this adoption, Neawedde said, and is the farthest behind in taking advantage of IOT digitization.  

The benefits of IOT are numerous, he said, including managing and minimizing risks. It can prevent plant downtimes, improve occupational safety, resolve warranty claims and avoid penalties for delays. IOT eliminates efficiencies by saving energy, reducing labor costs, using resources efficiently and optimizing along the value-added chain. 

It also allows a facility to optimize investments so that there are less redundancies, fewer spare parts and longer runtimes. User experience is improved with IOT offering a more satisfying way to do things, Neawedde said. It is easy to use and easy to share, and the configuration is flexible. IOT can solidify a company’s market position. 

“You’re doing something that is different and gives you a competitive advantage, with less risk and higher efficiency,” he said. “If you don’t do something, your competition probably will be.”

One way to utilize IOT is to build a digital powertrain for a key asset by adding smart sensors to all components such as motors, drives, pumps, couplings and gearboxes, and bearings. Information from those sensors goes to a gateway that pushes all that data to the cloud. 

Each component can be monitored individually or as part of the complete powertrain, Neawedde said. This provides transparency on how the different powertrain assets are working together, allows for process optimization and provides the potential for energy savings. Once in the cloud, analytics and dashboards can be created to monitor the health of the equipment in real time. 

On a daily basis, users can see what is having an impact on the lifespan of the equipment and the expected lifespan based on the health of the components. 

Actual aging of the equipment is provided, showing when to start planning for maintenance and when the equipment needs to be repaired. 

“Now I know exactly when the equipment needs to be repaired and the expected lifespan,” Neawedde said. “I can do this for individual assets, for a line or for the overall health of the system. If I want to, I can drill down and see which individual component is probably on the worst path.”

Company-wide monitoring is possible from one consolidated dashboard, he said, providing a roadmap for all assets. It’s possible to spot problematic assets, perform root cause analysis, anticipate faults, operational issues and trigger maintenance actions based on condition information. 

“I can compare plant A to plant B,” Neawedde said. “I can see what my maintenance budget is going to need to be two to three years ahead of time. You don’t want to do replacement for replacement’s sake. You want to replace the ones that make the most sense.”

The digital powertrain accomplishes all the goals of all the four maintenance strategies, he said. 

“You have time to react, to have a solid plan,” Neawedde said. “It’s going to save you a lot of money.”

IOT has improved significantly the last several years, in terms of a lower cost and the ability to interpret all the data that is being collected. For the first seven or eight years, no one knew what to do with all the data. 

“Now we’re at the stage where we’ve been collecting data and have the analytics to dig into the whole system and understand what’s going on,” Neawedde said. 

In the past, there were infrastructure issues because there wasn’t the fiber or the communication bandwidth to connect sensors to a central data collection point. With IOT, that is no longer a concern. 

“Narrow band IOT uses cellphone connectivity to go up to the Cloud, or you can use Bluetooth,” Neawedde said. “It’s independent of production so it doesn’t slow down what’s going on and it doesn’t interfere with safety. It’s collecting the data and pushing it to the Cloud or somewhere on premise.” 

Security concerns are minimal because the IOT device transmits but it doesn’t receive anything — it’s a one-way transmission, he said. 

Case studies 

Olam International, a global supplier of food and industrial raw materials, added 100 IOT sensors on motors in its cocoa facility in Singapore, a dairy processing plant in Malaysia and a sugar refinery in Central Java, Indonesia. 

In the past, motors were monitored manually, costing time and labor. The remote sensors allow for predictive maintenance, reducing downtime and extending equipment life, Neawedde said. The savings from preventing one motor failure has recovered Olam’s investment in the smart sensors.  

Pagen AB, a leading family-owned bakery in Sweden, added smart sensors and gateways, replacing handheld vibration measurements that were done four times a year. With the sensors, vibration measurements are updated one time per hour by remote connection.