The largest cause of disruptive and expensive unplanned downtime in manufacturing plants is motor breakdowns. Each production line has dozens, even hundreds of motors, and each one is vulnerable to unexpected problems. In large plants with multiple lines, the motor count can go into the thousands. Tiffany Huang, research analyst at Lux Research (photo, left), made the case for the value of next-generation sensors in Cleveland last week at the Advanced Design and Manufacturing show.
In the session, "The Next Generation of Intelligent Sensors for Better Visualization," Huang explained the benefits manufacturers are receiving from deployed systems of sensors that include data collection and analytics.
“Valuable ROI means cost reductions. The sensor can be deployed for predictive maintenance or safety for those working along with robots,” said Huang. “Also, sensors can give you a complete view of your facility, and that can mean cost reduction.”
Sensors for Maintenance and Safety Both
Part of the reason sensors are becoming a bigger part of the manufacturing mix is because they’ve come down in price significantly. “You now have sensors on your assets. They’re getting cheaper and smaller, and they’re going mainstream,” said Huang. “For maintenance, you want sensors that tell if something is about to fail. That means sensors that detect vibration, temperature, ultrasound, or increases in noise levels.”
Beyond predictive maintenance, sensors are also working to support quality control and to determine when an asset has reached the end of its useful life. “For efficiencies, you can have in-line quality control, imaging techniques, optical detection, or chemical detection,” said Huang. “Also, you can measure machines to see if they’ve been used to their maximum.”
Preventing worker injury is another value of new sensors. Next-gen sensors can detect dangerous situations. “For safety, you can place sensors on machines or on workers. In the self-driving space, you find sensors that detect collision. You can use these on workers so you know if they’re in danger,” said Huang. “Also, you can use sensors to let workers know if they’re in a hazardous environment. These sensors can tell the worker what garments to wear.”
Deploying an Effective Sensor System
A large part of a sensor solution is the ability to collect data, store it effectively, and then analyze it to produce effective action. “Connectivity has become easier. At the basic level, you can go with Ethernet. That’s good if you have a couple sensors. If you have than a few sensors, you can use wireless connectivity like Bluetooth or WiFi,” said Huang. “Longer range sensors may require cellular smartphone technology. There is also new, innovative, low-power wide-area connections.”
Many company now offer packaged services that include sensors and a connectivity system. The packages typically include the ability to store the data and perform analytics. “Unification is important to make sure all your data gets into one place. Analytics horsepower is important so you can get insights from the data,” said Huang. “The user interface also matters. You don’t want to have to train someone for a year. You want to make it as intuitive as possible. And access is critical. You need to make sure the data is available from several sites and accessible from different devices.”
Automotive Plant Saves $2 Million in Reduced Downtime
Huang cited the example of sensor system deployment where an automotive manufacturer wanted to reduce downtime related to maintenance issues. The company wanted to reduce downtime by 50%. The company deployed a system from Senseye, a preventive maintenance systems firm.
“They put sensors on 12,000 machines across three production lines,” said Huang. “They connected motors, gear boxes, and pumps for predative maintenance and diagnostics. They used sensors to detect vibration, velocity, and acceleration, sending all the data to historians for analysis.”
The Senseye package included sensors, data collection, and data analytics. The sensors were connected by Ethernet, and the system organized the collected data. “Senseye has machine-learning techniques that know what a healthy machine looks like versus an unhealthy machine,” said Huang. “They also predict upcoming health issues that might affect a machine within a couple months.” Huang noted that the system was able to catch health issues with an 80% accuracy. “They saved $2 million in the first year by reducing unplanned downtime.”
Rob Spiegel has covered automation and control for 17 years, 15 of them for Design News. Other topics he has covered include supply chain technology, alternative energy, and cyber security. For 10 years, he was owner and publisher of the food magazine Chile Pepper.