LAS VEGAS — While incorporating the internet of things (IoT) into baking plants may seem like a futuristic and daunting leap for industry executives, modest introductions of this rapidly growing technology into facility maintenance are simple and valuable, said Jay Wright, vice-president of sales and marketing, Somax, Inc., Alpharetta, Ga.

By incorporating interconnected sensors to monitor the temperature of baking equipment, bakers have the potential to save on labor costs and reduce the risk of damaging equipment and costly downtime, he said.

Mr. Wright spoke Sept. 7 at IBIE 2019 on how IoT will be incorporated into the maintenance function of the baking industry in coming years.

The case for IoT in baking maintenance has become increasingly compelling because of the lack of resources bakers have to devote to maintenance, excessive paperwork in maintenance, the increasingly complicated array of software involved in bakery production today and what Mr. Wright said was a lack of support for maintenance from management.

Even more important, he said, is the increasing shortage of skilled maintenance personnel available in the marketplace. He cited data indicating about 700,000 maintenance jobs open in the United States recently. Young people currently aren’t being trained in industrial arts courses for the multi-faceted work required of maintenance staff, and many of the most talented of these workers employed currently are rapidly approaching retirement age.

The challenge of rising labor costs prompted the baking industry to pursue production automation decades ago, Mr. Wright said, adding that automating maintenance operating is the next frontier.

He described IoT as:

“Interconnected devices usually going to the cloud or some central hub, gathering individual points of data going to the cloud or where it is processed. All that information can be brought back and something tangible or useful comes out of it,” generally to mobile devices.

By 2020, 20 billion devices globally will be connected by IoT, a figure he said is “just scratching the surface of its potential.”

The key to deriving value from IoT is not the gathering of the data, but generating information that is accessible, may be insightfully applied and shared throughout a company, he said. PLCs gather data, but it is isolated, Mr. Wright said.

As an example of how IoT may be used in baking, he suggested the installation of $100 sensors on a piece of baking equipment that will replace weekly or monthly inspections of the equipment. By contrast, the sensor will check on the device once a minute. Next is communication through gateways that gather data and push it to the cloud. He recommends a cellular gateway.

He noted that currently if a piece of equipment begins having trouble after an inspection, days if not weeks may pass before the problem is identified, leaving the device prone to damage. Continuous sensoring identifies a problem as it develops and alerts staff whether the staff is present at the plant or not.

Mr. Wright said baking is 5 to 10 years behind many other industries in adopting IoT for maintenance, a delay that has been beneficial to the real degree the cost of IoT has been declining.

While many baking operations managers identify vibration as a good test case for using sensors, Mr. Wright advocates heat sensors as a better starting point. He said detecting changes in vibrations is complex, while heat sensors are capable of identifying a wide range of issues, including low oil in a gear box; alignment issues; binding issues; loose or faulty wiring and bearing issues. Importantly, heat detection allows the baker to catch problems before significant damage occurs.

Success with IoT comes not just from identifying problems, Mr. Wright said. Developing checklists/procedures for what to do with the information is key.

To begin using IoT, he said a baker could buy a number of wireless temperature sensors and run them for one to two weeks to establish a baseline for normal temperature. Notification thresholds would then be set. For instance, at a 10% divergence from target, the maintenance crew would be altered to check the device for a number of potential problems. If the divergence grew, say to 15%, maintenance work would be done. If the divergence grew still further, management would be notified of impending downtime while the equipment and its line were shut down for maintenance.

The cost savings from implementing IoT may be extraordinary, Mr. Wright said. A maintenance worker may spend 30 minutes each week to inspect a machine. At $17.50 a week, the cost would come to about $910 per year. Using a sensor will cost about $150 the first year (including $100 for the sensor) and $50 per year afterward. IoT will not eliminate the need for maintenance staff but will reduce their numbers, Mr. Wright said.

Drilling deeper into the workings of IoT for maintenance, Mr. Wright said smartphones and tablets should be the device of choice for incorporating this technology.

“Smartphones are the new clipboards,” he said.

Reducing plant noise, by eliminating communications through intercoms or walkie-talkies is another plus.

Maintenance and sanitation workers should not be sent to personal computers for information or to input information, Mr. Wright said. The workers are the most highly paid at a plant and often are not talented when it comes to personal computers. IoT maintenance and sanitation documentation should be done on mobile devices.

Other examples for use of IoT in bakeries Mr. Wright cited ranged from sensors of moisture and humidity and air compressors (“nothing shuts down a plant faster than a lack of air compressors”) to rodent traps.

“Understand that you don’t need to do everything at once,” he said. “Don’t try to bite off too much. We can start with a slow, methodical approach. It’s a path to digital transformation. We have no choice. The people aren’t there.”

Over time, Mr. Wright said marrying IoT with big data will allow bakers to compare maintenance across multiple plants. Even further in the future, sensors will be able to predict problems and offer prescriptions. For example, a system could warn a plant “Mixer 3 has a bearing that will fail in three to four weeks. Do you want to order a replacement now?

“We have to adapt. How is the IoT changing maintenance? We’re making it possible to do it now. Right now we’re struggling. Plants I talk to have four or five openings for every shift. We have to get to a better way.”