Using connected sensors and machine learning, building operators can get advanced notice of potential systems failure and schedule on-time maintenance or repairs.
Occupancy sensors, weather conditions, usage statistics . . . all create a comprehensive view of equipment wear and usage. Pairing near real-time data with machine learning can help determine potential equipment failure well before it occurs.
The intelligent building needs to be sustainable
Today, most equipment has detailed instructions and procedures to perform preventive maintenance regularly. Many systems require operators to check for wear and tear, replace some components, clean filters, or otherwise make updates and adjustments at regular intervals.
While these procedures can help avoid or delay equipment failure, they don’t take into consideration many differences in the use of the devices or machines, the conditions of usage, and other factors.
Over the past few years, manufacturers of elevators, HVACs, escalators, security systems, and other equipment for buildings and building management have started featuring machines and components with sensors and remote-control capabilities.
New equipment with embedded sensors not only helps give a picture of the different statuses of the machines, but it can also help operators to predict when it is necessary to perform maintenance procedures to keep the units in working condition and safe.
Machine learning and artificial intelligence can help schedule different maintenance times for large installations
Preventive maintenance is typically based on time or usage. Predictive maintenance, on the other hand, is based on the actual condition of the equipment, not the calendar.
While predictive maintenance does not replace the compliance-driven requirements of routine checks, it eliminates what some facility managers dub the “clipboard” approach.
Data collected from several units can help manufacturers paint a picture of the usage conditions in different areas and environments. For example, an HVAC installation in a dry, hot area could work differently than another one in a cold and humid climate; the same model installed in a residential building works at different times and conditions than another in a factory or office building. Temperature and humidity sensors can provide a clear picture of those conditions.
In the example above, the typical maintenance manual of the HVAC usually doesn’t specify different usage conditions. It most likely indicates performing preventive maintenance based on the number of hours in operation or between seasons.
Elevators are another excellent example of the advantages of predictive maintenance. Real estate maintenance company Enertiv says the benefits are impressive, thanks to collected data. IoT predictive maintenance used by the company reduced maintenance costs by an estimated 25%, with a 50% reduction in equipment failures, and extended equipment life between 20 and 36%.
Knowing different conditions and using historical data, operators and service providers can analyze usage patterns and previous maintenance issues to determine the possibility of equipment failure in certain situations.
Another benefit of collecting sensor data for predictive maintenance is the ability to generate an auditable trail of machine performance and behavior that you can use to your advantage should you file a warranty claim.
Predictive maintenance can help optimize schedules and supply chain
Today, supply chain management is a must for most organizations. The shortage of materials, especially electronic equipment, can disrupt manufacturing facilities and the regular maintenance of existing operating units in the field.
That’s why knowing in advance when and where spare parts will be needed, could help keep things operating with little disruption in operation. Furthermore, spare parts can be requested in advance, and properly trained technicians can be scheduled to perform the necessary maintenance on time to avoid expensive off-hour repairs.
Investing in predictive maintenance pays off
Predictive maintenance is appealing across many industries. While the 4th industrial revolution – dubbed Industry 4.0 – is still far from being widely adopted, between advanced analytics, big data, edge computing, and the cloud, we have a good idea of what the industries of the future will look like, and how far manufacturers and service providers need to go to complete their transformational journey.
So far, for the building industry, there are clear benefits to adopting the new technologies:
- • Reducing catastrophic equipment failures: Using IoT sensors and powerful analytics, maintenance company Enertiv has experienced a 50% reduction in significant equipment failures.
- • Reduced maintenance costs: The facility manager of one 29-story office building saved more than $16,700 in operating expenses and another $32,300 in repair costs annually.
- • Better safety performance: As emergency repair jobs are inherently more hazardous for maintenance personnel and put more workers in harm’s way, predictive maintenance helps avoid those scenarios, improving safety.
- • Managing warranty claims: new equipment under warranty can fail for many reasons. Using the data collected and analytics helps facility managers file warranty claims with specific data, especially when failures are chronic and repetitive.
Industries such as heavy manufacturing, railways, energy production, and aviation are already significant users of predictive maintenance. An equipment malfunction can create massive operating disruption for those sectors and cost millions of dollars a day.
Looking ahead
The building industry is slowly waking up to seeing the advantages of IoT and artificial intelligence. As more intelligence is built around equipment and facilities, predictive maintenance is a way to improve operational efficiency and safety, as well as reduce costs.
