Sensors drive intelligence into maintenance

Data-driven analysis can create a roadmap to find problems early.

By Mario Calvo, Parker Hannifin Corporation August 4, 2016

It seems like everything we buy today is intelligent. Thermostats that adjust themselves based on weather conditions. Cars that alert us to upcoming maintenance needs. Alarm systems that can be armed or disarmed from a mobile phone. Wristbands that track fitness progress.

Thanks to the Internet of Things (IoT), the same technologies that have revolutionized the way consumers do everyday things can now enhance the way we manage maintenance in the plant environment. And it doesn’t take a brand new plant or the latest equipment to derive benefit.

The spectrum of maintenance ranges from passive to predictive. On the passive end, maintenance is approached primarily as a fix-a reaction to a problem, issue, or breakdown. When examining the potential disadvantages of this reactive approach-downtime, production delays, inefficiencies, added costs, poor quality, increased scrap, warranty issues, or product recalls-it’s easy to see why it’s not a best practice.

On the other end of the spectrum, predictive maintenance employs ongoing analysis to determine the condition of equipment and predict when maintenance will be needed. The more compelling selling points of this sophisticated approach are its role in preventing unexpected equipment failures as well as its ability to identify precisely when maintenance should be executed.

Somewhere in the middle is condition monitoring. A practice that is quickly becoming fundamental to maintenance management, it utilizes periodic assessments to analyze the health of equipment. When executed consistently, it enables an understanding of a more periodic state of the equipment-a heartbeat of its condition, if you will-to help inform maintenance planning.

Condition monitoring identifies trends that may indicate a change in the way equipment is performing. Such changes often serve as early warnings of impending issues. Recognizing them can be the key to triggering action that prevents failure-and its consequences.

It wasn’t that long ago that the only option for monitoring the condition of equipment was to do it manually. Anyone who has spent any time in a plant knows the process can be arduous, especially for operations with significant equipment inventory. Visually reviewing gauges, entering data into spreadsheets, traversing into hard-to-reach or dangerous areas-the traditional approach is time consuming and vulnerable to human error. And, oftentimes, it’s not proactive enough to prevent a fault or failure.

Sensors are making over this historically manual process. Deploying them to monitor different elements of equipment-pressure, temperature, humidity-can pinpoint conditions that may be indicative of a specific problem.

Sensing maintenance needs

Sensors are a cost-effective way to take a snapshot of what’s happening with a piece of equipment-down to the second, if that’s what an organization needs. The technology is fast and efficient, enabling the monitoring of more process points and more machines. And it’s suitable for use in a variety of applications. In fluid power, for example, sensors can be applied for condition monitoring of injection molding units, metal forming and fabrication equipment, conveyor systems, dispensing systems, robotic assembly, and hydraulic power units, to name a few.

In the fluid power example, sensors can assess systems that use oil, water, or gas to either transfer liquid or gas from one point to another or as a means to generate power. Pressure, temperature, and humidity are absolutely critical in these systems, especially when it comes to quality. Too much or too little pressure can wreak havoc on production. And if temperatures don’t remain consistent-in the tool, lines, and atmosphere-it can be to the detriment of quality. Sensors can provide enough data to enable the identification of trends that show the condition of a piece of equipment is changing. This insight can trigger corrective action, before the catastrophic occurs. Sensor-driven condition monitoring is especially useful in hard-to-reach or inaccessible areas where on-board gauges may be hard to reach or view. It’s also valuable in hazardous environments, enabling the user to keep a safe distance and eliminate the need for special safety-related clearance or process.

Because replacing manual monitoring with sensor technology enables the optimization of assets and processes, productivity is arguably its biggest benefactor. The first productivity gains come with the elimination of the need for staff to perform manual reads of equipment conditions, which can net a significant improvement in labor efficiency. The second gains, and likely even more significant, come with the prevention of downtime.

According to the U.S. Department of Energy, minimizing downtime has been shown to drive up asset availability by as much as 30% as well as increase production by as much as 25%. By identifying issues before they escalate, organizations can also decrease maintenance costs and avoid potentially dangerous situations.

Going wireless

When organizations learn about how sensors can be applied for maintenance management and condition monitoring, they can’t deny the benefits. But the added-on, and often high, investment in the installation and use of traditional wired sensors can make implementation cost-prohibitive.

Wireless technology is a viable alternative. Wireless options make sensor solutions surprisingly affordable. As consumer technologies have pushed wireless into the mainstream, it has driven down the cost of the bill of materials needed for wireless sensor acquisition, especially components like Bluetooth radios. And because wireless doesn’t require strenuous installation or specialized skills, implementation is more affordable.

The accessibility and ease of implementation makes wireless sensors ideal for placement in many types of production machines to monitor a variety of operating parameters. Reduced cost isn’t the only benefit of going wireless. The sensing technology is easier to install than its stringing wired counterparts, and it eliminates risk of physical connection faults. The addition of access points is simplified, and the commissioning routine doesn’t require laborious programming and data-collection devices, such as a programmable logic controller (PLC). Once installed, a wireless sensor allows for instant and easy access to data for a variety of stakeholders via user-friendly software and apps.

Enhancing equipment—new and old

Some of today’s newest equipment comes with on-board sensor technology. But it’s safe to assume that older equipment-which still accounts for the vast majority of the assets found in plants around the world today-doesn’t have the same bells and whistles.

It doesn’t take a new asset to implement sensor-based condition monitoring. Sensor technology is so flexible and adaptable that it can be implemented for maintenance management throughout the lifespan of equipment. Even for new technology, organizations can consider aftermarket application of additional test nodes to increase their monitoring potential.

Wireless condition monitoring technology can readily supplement an existing wired system without being directly integrated into it. More importantly, cost-effective wireless sensors provide users with the flexibility to add more nodes to the process points that make the most sense to their specific applications-a capability not available to industrial organizations as little as three years ago.

Putting data to work The collection of data through sensor technology is just the beginning. The data mean nothing if they aren’t utilized.

The other critical element of a sensor-based condition monitoring system is the software application that houses and provides access to the data. These applications make information available on multiple devices. And they can be configured to generate dashboards that transform data into insights that drive maintenance management decisions. Additional support for decision making can be provided by features such as alarms or e-mail alerts that notify users of certain conditions in real time and allow them to act.

Taking intelligence to the next level

Effective, disciplined, condition monitoring lays the groundwork for-and is the precursor to-predictive maintenance. In fact, the data generated through condition monitoring are mission-critical for a predictive approach. For organizations that aspire to move along the maintenance management spectrum, a sensor-driven condition monitoring program can be part of a deliberate roadmap to achieving predictive maintenance.

In the near term, the options for how condition monitoring data can be used are becoming more robust. For example, organizations may be gathering maintenance data from a variety of systems. But it’s likely those systems aren’t interconnected and their dashboards are exclusive. Incremental improvements in technology will help marry data sets. An organization can overlay what’s happening with its hydraulic equipment, for instance, with weather data or information about the utilization of energy. This enables the optimization of various factors that can affect production, such as ambient temperature conditions of the production floor that impact production equipment or energy-saving discounts offered by utility suppliers when energy utilization during peak times can be minimized and validated.

It’s possible to make maintenance management as intelligent as some of the consumer technology that’s become so commonplace today. Sensor technology is the powerhouse that can make it happen.

– Mario Calvo holds three patents and has been with Parker Hannifin Corporation for 22 years, currently as business unit manager for its Diversified Technologies group focusing on life sciences and IoT.