Is IIoT the future of maintenance?
By James Reyes-Picknell
By James Reyes-Picknell
Plant and equipment monitoring is very basic to most maintenance programs, even if it is not done by maintainers. Condition monitoring (CM), as it is known, can account for a third or more of your maintenance program budget, if you are taking full advantage of it. You can do it with various hand-held or permanently installed technologies, or with the human senses. Often it is operators who apply their senses doing rounds, and noticing when things are not quite right. They may not know what is wrong, but they know something is amiss. They contact maintenance, who typically responds with a technician and some hand held monitoring tools. When you’ve “diagnosed” the problem, you can plan a course of action – job plan, arrange parts and other resources, and schedule the downtime when it convenient.
It’s that ability to act in our own time, not when the machine breaks down under load, that gives us value from CM. Waiting for breakdowns is expensive and risky. Acting proactively eliminates much of the risk and enables us to avoid major business losses.
The “new game” in town is the Industrial Internet of Things (IIoT), which is already making big splashes in the world of CM. Here’s a look at the impact it will have on maintenance in the not too
For a long time, we have had sensors for motion, ultrasound, light, temperature, vibration and more. Some are installed in or on equipment, and wired to monitoring analyzers, while others are carried around, and moved from machine to machine. For decades, these have been helping companies to detect failures in early stages, so that they take action to minimize the consequences of the failures they are detecting. Hand-held stethoscopes, screwdrivers held onto bearing housings, and hands-on equipment, have been replaced with vibration sensors, temperature probes, ultrasonic and thermal imaging.
Some equipment comes with continuous monitoring; diesel generator sets or packaged compressor units have temperature, vibration, oil pressure, and other monitors installed, all sending signals back to a monitoring panel. Someone had to monitor the panel – often maintainers.
CM technicians can perform rounds and monitor other equipment, at least periodically. Companies will do that for you. Many of us have had a great deal of success with this – with many “saves,” but some were missed. This monitoring, if done by a contractor, is often too infrequent, but even with the misses, the saves can make it worthwhile.
CM is of great value in giving advanced warning of failures about to happen. The more frequently you monitor, the more likely you will be to catch failures. Monitoring a vibration signal once a month is not enough to catch all failures of rotating equipment that move through a period of degradation within weeks. Installing permanent sensors enables continuous monitoring, but it was very expensive. You would need to install the devices, wire them, and set up a monitoring station. The biggest cost was often the wiring. Someone needed training in the monitoring technology to interpret the signals and makes the judgement calls about what is a failure. Companies will do continuous online monitoring for you, and tell you when a problem is detected. Using them you can have your equipment monitored from anywhere. This may be less costly than training your own technicians, but will limit the list of equipment monitored to only critical production assets.
Because of cost, these installations have been used by the largest companies. They can afford the contract monitoring or train their own technicians. Smaller operations with thinner margins could not. They lived with (and live with) the risk of surprise failures and consequences of failures. They can’t afford those failures, and a big one could put them out of business, but they really had no option that they deemed viable. Smaller organizations often lacked technical expertise to examine these options.
Now IIoT has changed that. It is removing much of the cost barrier to companies that couldn’t afford the permanently installed hardware, wiring, software, and training of their technicians. It is also opening up more than just critical major machinery to continuous monitoring.
IIoT is an electronic industrial network of connected devices monitoring your equipment, processes, and systems. Some devices talk to analyzing equipment, while others talk to each other. Most of them are wireless. Transmitting signals via Bluetooth, LTE, or local area Wi-Fi networks. Many machine mounted devices also have the ability to monitor more than one signal (vibration, temperature, and sound levels). Many have some computational capability and machine learning built in. They “learn” what normal is for their application and use it as a benchmark for comparison, monitoring continuously. Once installed and activated, they spend time monitoring and learning what a normal signal should be, and are then ready to detect when abnormal signals occur.
That onboard computational capability is known as “edge computing.” It removes a lot of load from the available bandwidth on networks. They monitor continuously and send data only when there is an abnormal condition. Without that capability, Bluetooth, LTE and Wi-Fi networks would be overwhelmed by the steady stream of data from a large and likely growing population of devices, each one monitoring some important parameter. It’s really only the anomalies that we are interested in though, so why monitor the monitor? Let it tell us when there is a problem, and then investigate with more analyzer capability or even visits to the field to see what is happening. Our brains are reserved for the tough problems, the machines do the rest.
While the sensors themselves must still be installed, the wiring for them does not. LTE and Wi-fi may already exist in the field and if not, are easily installed. Bluetooth can be monitored by mobile monitors, handheld tablets carried by plant operators or maintainers. One day, perhaps even drones programmed to fly through the plants can be used to carry out “rounds.” The days of the clipboard are almost over, perhaps even the rounds person won’t be needed soon.
Behind the scenes we don’t really need to monitor all the signals all the time. Edge computing does that, but someone must receive and act on the warnings generated by abnormal signals. Those abnormalities might be device faults, network faults, or machinery defects that are being detected.
When we get those alarms, we must act – if not, we won’t realize the benefits. For the vast majority of defects, we know we will be stripping equipment apart and replacing parts, then putting it back together. For many cases, we don’t need in-depth analysis of what is wrong. A simple field visit by a technician might be enough to tell us what repair job we need to prepare for.
In other cases we might be monitoring major equipment that has the potential to take our operations down. Downtime to repair may be long, so we need to know more about the nature of the failure. Once an abnormality triggers an alarm, the devices can send a complete stream of signal output to a monitor that is being operated by a technician or engineer, who views the live stream of data, performs analysis, and pinpoints the source of the problem. Parts can be ordered, job plans prepared, and crews readied for the work, so downtime is minimized.
Some equipment might be remote or difficult to access for a human technicians. The IIoT devices allow for monitoring without human visitation. This can be a big benefit in cooling towers, oil or water fields, remote pumping stations, wind turbines, and other unmanned operating locations.
There are several advantages to IIoT over older CM methods. The cost to install is much lower, and can be installed anywhere that has network coverage. The cost of devices is coming down. As more devices are sold and manufactured, their costs and pricing goes down too. What would cost millions to monitor in the not too distant past, can be monitored now for thousands.
With continuous monitoring of various signals a wide range of potential problems are picked up early. With machine learning, devices’ “signatures” are known and with AI, abnormalities in those signals can reveal specific problems. As companies permit sharing of their monitoring signal data, the ability of machine learning to expand and apply AI on problems. Data gathered from equipment in many locations can all be used to help solve problems. Until then however, we need a
Human monitoring of these signals is reduced only to the most complex and likely unusual problems. As this capability expands, the cost of monitoring will also come down. The location of the technicians doing the monitoring can be anywhere, not necessarily in your own head office, own country or even continent.
IIoT enables us to take full advantage of our connected world, to use talent and capabilities that exist elsewhere to solve real-time problems, here and now. Furthermore, it allows us to do all that a fraction of the costs incurred for this kind of attentiveness just a few years ago.
Monitoring of equipment can expand. As prices come down and the benefits amass, we can monitor a broader range of equipment. Where it wasn’t “worth it” in the past, it becomes more viable. Maintenance programs become more biased towards CM. Reliance on preventive replacements and “inspect then repair as required” will be reduced. We shift further into the proactive realm and further away from “break then fix.”
As that expands and device pricing goes down, its capability is going up. A broader range of smaller companies with smaller margins can now afford it. As they become part of the IIoT market they benefit from being more proactive and less vulnerable to impacts of those big unexpected failures. Their business becomes more stable and more profitable.
Maintainers themselves are freed from mundane monitoring and inspecting tasks and freed up to what they really trained for and love doing. Fewer are needed for the CM. With fewer of the really bad breakdowns there will be less of the more difficult and riskier repair work. The work we have can be planned and scheduled. Safety performance of maintainers can increase. Productivity increases and makes it easier to maintain with the fewer technicians we have today. For operations in remote locations that probably means fewer people on-site and traveling back and forth. It’s a good fit with the practices were seeing to ensure healthy working conditions as highlighted by the 2020 COVID-19 crisis.
The CM industry will also shift. It will still exist and move to monitoring online with staffs of technicians, rather than selling devices to end-user technicians in the field. Contract monitoring services will likely shift off-shore to locations with well-educated yet less-expensive work forces.
IIoT is a game changer that will have a big impact, one device at a time. MRO
James Reyes-Picknell, PEng is Principal Consultant of Conscious Asset, providing business consulting and training services in Physical Asset / Maintenance Management and Reliability. He is author of several books, including Reliability Centered Maintenance – Reengineered in 2017 and Uptime – Strategies for Excellence in Maintenance Management, 2015.