The latest, smartest generation of sensors costs up to 10 times less than the sensors of five years ago, enables shorter cycle times, screens out interference and can be reconfigured automatically. The dramatic drop in price makes it cost effective to add sensors to less expensive, previously silent equipment. And harvesting this newly available data is changing the face of manufacturing.
“People want connectedness—the ability to see what is going on not only at the machine level but down to the sensors—to know exactly what is going on in the process,” Henry Menke, product marketing director at Balluff, said.
Other factors that make a difference: the digitization/shrinking of technology, widespread use of smartphones with Bluetooth, better control systems, and improvements in the I/O link communication protocol that communicates with sensors. In some cases, there’s even an app for that.
“With the advancement and digitization of technology, we’re able to embed brains into smaller and smaller devices at the sensor level,” Jim Gilbert, product manager of photoelectric sensors at Sick Inc. said.
“In the industrial world, we’re able to use advancements that make smartphones smarter and better. Technology that was the size of a deck of cards is now smaller than your little finger. Scales of economy have led to cost-effective communications in products that were previously silent,” he added. “We’re able to add intelligence to the sensor itself. What previously was limited to products with a value of $10,000 or more, now you can put in things that cost less than $500.”
Instead of indicating only whether a machine is on or off, Sick’s sensors run with a program that indicates if a machine is running out of spec, Gilbert said. “We can identify emerging maintenance issues. We can do remote diagnostics from hundreds of miles away and help end users understand, ‘This item needs to be serviced’.”
Other manufacturers have seen similar gains from designing technology that is both smaller and less expensive.
At about $300 per sensor, ABB developed its newest sensors to be cost effective, Kyle Davis, smart sensor technical support engineer at the company, said.
At that price point, it can make sense to use ABB’s smart sensors throughout a manufacturing plant.
“In the past, it was worth putting $3000 in sensor monitoring equipment on your $100,000 machine because obviously you would want to know before something catastrophic happened to the expensive machine,” Davis said. “But people said, ‘We’re not going to put $3000 in monitoring equipment on a $6000 low-voltage motor. We’ll just run these until they fail’.”
When the smaller motors did indeed fail, production halted until they could be repaired.
“Just because it’s a low-voltage motor doesn’t mean it can’t completely stop production,” he said. “Now, with more cost-effective technology, you can perform predictive maintenance and prevent that middle-of-the-shift failure.”
Instead of stopping production and using a probe to check on a machine, workers using a smart sensor app from ABB can monitor the condition of a motor and its bearings using a smartphone and Bluetooth—meaning no new structure or wires are needed, Davis said.
“Through Bluetooth technology, we can communicate back and forth between the smart device and the sensor itself using a smartphone app,” he said. “Rather than having to install all this structure and run wires, it’s all wireless. Everybody walks around with some kind of smartphone now. We capitalize on what everybody already has. The only thing you pay for is the actual sensor.”
Predictive maintenance based on data from the sensors helps reduce downtime by up to 70%, extends motor lifetime as much as 30% and reduces energy consumption by up to 10%, Davis said.
Bad bearings cause up to 60% of all low-voltage motor failures, he noted.
“The system looks at multiple parameters to understand what the bearings are doing and makes a cumulative assessment of what’s going on,” Davis said. “You reduce downtime because the motor is not failing at midnight on Saturday when you’re going to have a hard time getting a replacement. Now, instead of waiting for it to fail, you plan for it and replace it during scheduled downtime.”
Because of the two-way communication Bluetooth affords with the sensor, sensors can be directed to monitor a specific motor.
The cost of ABB’s new sensors includes the sensor, the hardware to mount it and access for two years to the app and web portal, he said. The intention is that, just like for smartphones, the technology will improve in two years and customers will replace the current version with the latest iteration.
ABB began selling its new sensors in Europe 11 months ago and completed the UL safe area certification a few months later, It expects to be able to provide UL-safe certified sensors by this summer. It also expects to get Division 2 certification for hazardous areas this year. Interest has come from cement and aggregate makers, open paper, food and beverage facilities and many others, Davis said.
Other sensor improvements have enabled manufacturers to become more efficient and to address issues that have slowed manufacturing.
Consider a maintenance worker on the factory floor. Reflected light from his reflective vest used in the past could trick a photoelectric sensor into mistakenly concluding that a box was still present.
“The sensor is meant to detect when a box moves across and breaks the beam,” Gilbert said. “A reflective vest can give a false signal to sensors in a production environment. Someone could be standing five or 10′ (1.5–3 m) behind the beam and have a reflective vest on. Even though a box is breaking the beam, the vest would send a false signal back to the sensor and the sensor would think there is not a box there. We call it ghosting.”
The first challenge was finding out what caused the ghosting. “In some cases, it took a long time to figure out the root cause,” he said.
The second challenge was improving the technology to mitigate or prevent the ghosting: Sick is now on its fourth generation of sensor technology with advanced algorithms to address the issue, Gilbert said.
“Now, sensors are able to distinguish between bright lights and reflections in the room,” he said.
Maintenance workers can come much closer to the production line without causing issues.
Thanks to other improvements, maintenance workers can focus more on other tasks.
Sick’s IQG inductive sensors, introduced in 2016, are one example of sensors that no longer need to be adjusted during replacement. Manufacturers can pause a production line, instead of completely shutting it down, and make such changes as “hot swaps,” Gilbert said.
“Now you can configure sensors with the push of a button on a PLC,” he said. “If one sensor becomes damaged, you can remove that sensor and replace it with a sensor with the same part number. The sensor will automatically download the parameters so that a maintenance person doesn’t have to reset the sensor.”
The latest version of IO-Link, a protocol that allows sensors to communicate with a PLC, converting analog to digital at the sensor itself, also has helped drive progress.
IO-Link 1.1 enables higher bandwidth than the previous version. And that means more data, such as process and configuration data, can be more easily added.
It also allows users to set multiple parameters to track sensors as opposed to tracking only if a sensor is on or off, Balluff’s Menke said.
Multiple sensors also can be programmed to respond quickly to batch-size changes in manufacturing, as opposed to individually changing each one, Gilbert said. “At the push of a button, the system is able to respond.”
Speaking of multiple sensors, Balluff now sells a photoelectric sensor, the BOS 21M ADCAP, that can be configured four different ways. Call it the utility infielder of sensors.
“Traditionally, we design a sensor to do one job,” Menke said. “This sensor does multiple jobs equally well. We have one part number that you can configure into four different sensors. This sensor allows manufacturers to reduce inventory.”
A manufacturer running the same machines can configure the sensors differently for different processes, he said. Or, a manufacturer might use the same sensor throughout the plant, performing different tasks on different machines.
“This sensor puts power in the hands of the users to query the sensors as they see fit,” Menke said. “We designed a sensor to have flexibility in its mode and what it can engage. Not every customer needs every feature, but the sensor has enough features to satisfy everybody.”
In one application, the sensor might suppress background noise or light. In another application, it might need to switch to a different sensing mode to avoid being tripped by a shiny target object, he said. The sensors, released late last year, cost just 10–15% more than other sensors that have only singular functions.
While running, the sensor can send diagnostics to indicate if its voltage, temperature and other parameters are out of spec, as well as indicate the amount of light the sensor is putting out and receiving and even whether it is dirty, Menke said. If a sensor fails, it can be replaced, and its replacement can be reconfigured remotely with the correct parameters.
Another advantage is that the sensor can work either on or off a network, enabling manufacturers to add the sensors now even if they do not yet have the network communication capability—enabling forward compatibility.
“This sensor is indicative of the move toward smart sensors to enable smart manufacturing,” Menke said.
Since the sensor converts a signal from analog to digital at the edge, at the individual sensor, it eliminates or lessens the need for the long, shielded cables required when communicating with analog devices.
“In the past, we would recommend to customers that they add a linear positioning sensor to their machines to address a particular problem and the customers would respond: ‘I’m not putting analog on my machine’,” he said. “They would rather live with their problem because they hated analog more than their problem. With IO-Link as an option, all the frustrations from analog go away. People can start integrating measurement devices easily into their equipment.”
Improved controllers mean manufacturers and shipping facilities now can control the exact portion of movement that matters most. Improved position sensors save time in production and shipping in automated factories.
Consider a robot arm that is picking up a package or piece of equipment. If the arm is moving too quickly, it will damage what it picks up.
In the past, that arm would run at a slower speed the entire distance.
Better technology means a sensor can accurately sense when the robot arm is within inches of what it will pick up, Gilbert said. The arm can be programmed to slow down only as it nears its target object.
“A hydraulic valve stroke might go from fully closed to fully open,” Menke said. “But when the process is running in a refinery, you might want to modulate around a very narrow range of 40% open to 43% open.”
The improved technology also lets manufacturers capture and analyze more data both in the cloud and at the edge. Each approach offers distinct benefits.
“We have hundreds of thousands of sensors in buildings that previously were really quiet—not much communication at all,” Gilbert said. “Now we’re pulling a ton more data. Control is extending farther from the PLC with onboard computing enabling decisions to be made at the edge.”
In the cloud “analysis turns all this data into very easily available graphics so that you can see a whole trend line of what’s going on,” Davis said. “You have a whole server’s worth of computing power.”
By monitoring long-term trends, operators can see, for example, when a bearing starts to wear down because it has started to perform differently, he said.
One customer saw immediate benefits from sensor-enabled, low-voltage machines, Davis said.
The company already has a service provider who comes by monthly to check the large machines. Now when they’re on site, they may also get a notification that one of the smaller, low-voltage, machines—now outfitted with sensors—is running outside its parameters.
“While they’re already walking around the plant, they can also pull and analyze the data for the low-voltage motors—it takes about 30 seconds,” he said. “They’re getting extra benefits from the same guys on the same routes who are already there. The corporate reliability guys like it because they’re not at every facility. They can look at all the motors anywhere in the world from their offices.”
On the other hand, operators on the factory floor also can harvest real-time information directly from the sensors.
“In the early days, people would try to do some of this analysis back at the PLC level,” Gilbert said. “By putting intelligence into the device itself, you’re able to make decisions more quickly. Things further and further away from the PLC can now be part of the information-gathering platform. Previously they were not.”
“When you’re right there at the sensor with a smart device, you can see what just happened in the last go-round,” Davis said. “Manufacturers are amazed that they can get the data at that price point.”