How IoT Devices Enable Predictive Maintenance

05/17/2018 | By Janelle Penny

Preventive maintenance is a great way to keep your equipment in shape, but the calendar-based strategy is a one-size-fits-all solution to a facility full of equipment with different ages, conditions and expected life.

Predictive maintenance, which emphasizes tune-ups and filter changes whenever each piece of equipment needs the work, can save time and money.

With Internet of Things devices keeping an eye on asset performance, it’s easier than ever. Investigate how a predictive maintenance model incorporating IoT technology could change the way you deliver facility management services.

How to Use IoT Technology for Performance Monitoring

Your Internet of Things-enabled monitoring system typically includes three components, according to Logan Soya, Founder and CEO of Aquicore, an asset management software platform:

  • Sensors will collect measurements from equipment and record ambient conditions.
  • A data logging or networking device will intercept the streams of data from of the sensors and aggregate them in a central location.
  • A dashboard or other interface will interpret the data and help you prioritize what to act on first.

“In terms of the data we can collect, it can be electricity consumption, water, gas, steam or BTUs to measure the energy the building is using,” Soya explains. “You can also measure utilities being consumed by a room or by a specific tenant.”

There are also a number of conditions for your equipment that can be monitored with IoT devices. For example:

  • Is a piece of equipment on or off?
  • How fast or hard is a drive running?

On the environmental side, Soya suggests using the Internet of Things to measure temperature, carbon dioxide or particulate matter. “All of those things combined create a well-rounded understanding of what’s going on in the building.”

Where to Start with Internet of Things Devices

Don’t be tempted to roll out in-depth monitoring for every building system right away. James McPhail, CEO of Zen Ecosystems, an energy management platform provider, recommends starting with one source of data, such as a smart thermostat, and building your Internet of Things arsenal from there.

“A thermostat can provide ambient temperature for the space and runtime data to show you how much the thermostat has to run to keep up with setpoints,” McPhail says. “The way to get the most out of the [Internet of Things] space is by adding other components – for instance, a thermostat plus a return air sensor. Then go a step further and incorporate real-time energy monitoring devices as well.”

Utilizing Internet of Things devices lets you eliminate the requirement for manual analysis, where you’d track degradation in return air, ambient conditions, setpoints and runtime separately.

Adding energy consumption monitoring and using a system that aggregates the data into one source of information lets you draw conclusions about failing compressors or motors much sooner.

On topic: Do your IoT devices risk a security breach?

After you’ve settled on which Internet of Things system to start with, Dan Burns, technology strategy director for real estate management company JLL’s Smart Building Program, suggests starting with an equipment inventory.

You may already have an existing list of assets in some form, perhaps as part of an existing preventive maintenance strategy. Then determine what qualities you’ll track as part of your switch to predictive maintenance and performance tracking. For example, if you’re starting with the HVAC system, you might want to track airflow, outside temperatures and vibration analysis for each piece of equipment.

Related: The 4 C's of HAVC Systems

“See what data you care about collecting first and then improve your building to collect more over time and refine the performance of your building,” Soya adds. “You might prioritize utility consumption first just because it’s the biggest bang for your buck. You can get a lot of savings from just monitoring one or two points of data, and it also acts like a heart rate monitor to help you detect failures. You might not know where the failure is, but you can know one is occurring. That’s an affordable place to start.

“Once you’re advancing your initiative and collecting other data, ask yourself about the business questions you need to answer, like ‘Why am I off from my budget?’ or ‘Why are my maintenance and repair fees skyrocketing?’” Soya adds. “Use those questions to drive what data you’re prioritizing. People tend to create a big lake of data and get lost in it.”

Switching to a Predictive Maintenance Model with IoT

As you gradually expand your Internet of Things data collection efforts to cover more of your equipment, start phasing out calendar-based preventive maintenance in favor of a predictive maintenance model that prioritizes maintenance where it’s needed most.

“We typically recommend that customers realign their maintenance schedules when they set up a management system. When the thermostats, sensors and energy monitoring devices are being installed, use that time to go through and do maintenance. Then you’ll start with a clean baseline and you can monitor against that,” McPhail advises.

Handle tasks that typically have a time-based expiration one of two ways, McPhail says:

  • Some customers take that information and then call their maintenance contractors.
  • Others share the alerts so that the contractors themselves receive the alerts saying “Here are the filters that need to be changed within the next 30 days” or “This unit is showing indicators of a low Freon charge or motor failure.”

For your mechanical system, Burns suggests enlisting the manufacturer or a third-party company that services multiple brands to conduct an initial vibration analysis survey. Other factors are easier to measure and interpret, including:

  • kWh consumption, kW demand and runtime: “If you see all of them go up, that’s a pretty good sign that a failure is going to happen. If only runtime increases, that might be more of a maintenance issue, like there’s a filter that needs to be changed,” McPhail explains. “If kWh increases but kW stays the same, that could be indicative of a fan belt going bad on a motor.”

  • Multistage equipment engagement: You should be able to see how long relays are engaged on multistage equipment, McPhail says. “If the second stage kicks in at the same time the first stage does, that’s terrible inefficiency,” McPhail adds. “If the second stage never comes on and the air conditioning is running nonstop on the single stage and never keeping up with the temperature, that’s a clear indication the compressor has failed.”

  • Frequency of use and overcompensation: “You can see if the motor’s turning over more frequently than it used to or if power factor has become an issue,” Soya says. “You might see a particular routine suddenly having to overcompensate because something within the equipment has broken down.”

Burns also recommends correlating individual equipment data with weather and building occupancy throughout the day so that if a building system is working harder, you can see whether the cause is more people in the building or a malfunctioning part.

Read also: Pittsburgh 2030 District Helps Create a More Sustainable City

“When you step back and look at the whole portfolio, you might say that because of humidity, temperature differences or all of the additional people in your building, the HVAC system might be strained today. The doors are opening and closing more, so conditioned air is escaping,” Burns says. “Monitoring environmental conditions allows you to have a more holistic view of how your building is operating and adds building-level context to your equipment. Then you can decide, do I have something that’s on the road to failure or have the characteristics in my building just changed today?”


Here are two hand-picked articles to read next:


Related Coverage