StartseiteBlogBlogCreating Predictive Maintenance Programs with Big Data Intelligence

Creating Predictive Maintenance Programs with Big Data Intelligence

Creating Predictive Maintenance Programs with Big Data IntelligenceThe two largest expense line items on a facility’s operating budget are typically utilities and maintenance. Facility managers of office buildings, manufacturing plants, retail buildings, schools, hospitals or any other large premises, are ultimately charged with reducing both.

In the FMLink chart below, each bar on the chart represents a site’s annual maintenance cost on a factored gross square-foot (FGSF) basis. The median cost is $1.23 per FGSF. Seventy-five percent of sites have reported a maintenance cost of between $0.70 per FGSF and $2.00 per FGSF. Taking the median of $1.23 per FGSF and multiplying it by the enormous spaces occupied by some manufacturers, retailers and other facilities, we begin to understand the scope of the problem; and the extent of the opportunity.

Typically, budget crunches lead facility managers to defer maintenance that is nonessential. According toFacilitiesNet, however, “Generally, a policy of continued deferred maintenance may result in higher costs, asset failure, and in some cases, safety, health and environmental implications.”

Reducing Costs While Improving Maintenance

Saving on the maintenance line item, however, does not have to come at a cost of asset failure or the environmental. In fact, there is a way to significantly reduce maintenance expenses while ensuring that equipment and machinery work even more efficiently.

When we take advantage of big data analytics and apply it to the energy consumption of facilities, we achieve results that were impossible several years ago:

  1. We move from resource-intensive preventative maintenance to leaner predictive maintenance
  2. We enable faster, better decision making that results in operational efficiencies and overall equipment effectiveness
  3. We significantly reduce energy consumption (and save costs)

Data-Backed Predictive Maintenance

When it comes to machinery and equipment, there are basically three types of maintenance schedules: corrective (reactive), preventative, and predictive. Corrective maintenance is the “wait till it breaks” approach; the “plan not to plan.” Unfortunately, it’s often the structure (or lack of structure) adopted in many facilities. According to MA CMMS, “Relying on a reactive maintenance is like flying blind, however, reactive maintenance is still the predominant way of working in North America. Approximately 55% of maintenance activities in the average facility are still reactive.”

As facilities grow and become more organized and budget conscious, they typically switch to a preventative-maintenance schedule. These approaches are based on time or machine run time and deliver procedures that are designed to detect, preclude, or mitigate degradation of a system (or its components). An effective preventive-maintenance program helps deliver savings of as much as 12% to 18% on average.

When a facility discovers its need for efficiency as both a cost-saving and sustainable best practice, it moves to predictive maintenance and realizes approximately 12% more savings over a preventative approach. By tracking and monitoring the energy profiles of equipment and machinery, we use the aggregated energy data to predict equipment failures and service only the devices that require service.

With predictive maintenance, companies are alerted to impending equipment failure. By eliminating unnecessary scheduled preventative maintenance performed on equipment that is not in need of service, companies reduce the maintenance expense and eliminate resource-intensive down time.

Decision Making and Operational Efficiencies

The systems that track device-level energy consumption enable many operational efficiencies beyond predictive maintenance. In a case study of The North Face’s energy management, the retailer installed a circuit-level energy-management solution in four locations. It clamped wireless self-powered sensors on the outgoing electrical wire from the circuit breaker on HVAC and lighting components in each store.

The results were phenomenal:

  • It discovered an AC fan system that was not working properly. Through early detection, the store management identified 69,420 kWh/year ($10,500) worth of savings and avoided an equipment failure.
  • The HVAC system in another location was not operating correctly, and air handlers were over cycling. Through the early detection, the store was able to save 16,016 KW.
  • Through real-time monitoring in San Francisco, managers got an alert on Mother’s Day that the store’s security cameras were down on that busy retail shopping day.
  • One store changed lighting schedule during off hours and realized a 10% annual energy savings.

Other companies find operational inefficiencies in the form of unmonitored BMS overrides, discovering unknown anomalies such as device idling and behavioral change in the corporate culture. Also, with ongoing commissioning and benchmarking of locations against each other, waste and other discrepancies are easily noticed and corrected.

Reducing Energy Consumption and Enabling Sustainability

Big data that is gathered from devices and systems, then aggregated to reveal trends, profiles, inefficiencies, benchmarks and maintenance alerts, has a welcome side effect: saving energy (and money) and enabling sustainability.

When anomalies are tracked and serviced as needed (predictively), and operational inefficiencies are corrected, companies are able to optimize energy use, increase production and improve processes. They save money on both the maintenance and the utility line items.

Big Data in the Energy Ecosystem

A data-driven approach is revolutionizing many ecosystems. From streaming music services and satellite navigation systems to pedometers, calorie counters, and heart rate monitors, we are all benefitting from the ease with which enormous data sets can be processed. In the energy ecosystem, big data enables us to understand and optimize the consumption patterns of equipment which leads to cost reduction and enables green operations of equipment, systems, facilities and businesses.

Originally published on The Data Center Journal