Key article highlights:

  • 1. Compared to reactive strategies, predictive maintenance can:
    • – Generate savings up to 40%
    • – Improve equipment productivity by 25% and reduce downtime by 45%
    • – Prevent and reduce catastrophic breakdowns up to 75%
  • 2. Retrofit solutions that enable condition monitoring significantly reduce initial investments typically needed for predictive strategies
  • 3. Latest approaches to Predictive Maintenance include:
    • Reliability Centered Maintenance: It prioritizes high-impact equipment for predictive maintenance while allowing the low-impact counterparts to be managed with reactive maintenance.
    • Dissimilarity-based Approach: A new model of predictive fault analysis that looks at patterns of differences rather than comparing individual machine behaviour to a standard baseline.

There’s currently an increasing focus being paid to energy-saving technology in the hospitality industry. This led to the predictable emergence of green hotel technology, but with profit margins remaining so tight, it’s impossible to not weigh them against their return on investments (ROI). Fortunately, efficient maintenance can help reduce overall costs with up to 10 times return on its initial investment. And if you’re not utilizing predictive maintenance, you’re wasting money.

Types of Maintenance

Whether it’s individual A/C units or a large scale chiller system, ensuring proper maintenance of essential HVAC systems is a necessity for successful hotels. Regardless of how you do business, you will utilize at least one of the following:

  • Reactive maintenance: Also known as “breakdown maintenance,” proponents of this style take no efforts to maintain a system and run it until it breaks down. Unfortunately, this one of the most commonly used mode of maintenance in most facilities.
  • Preventative maintenance: Performing time-based routine maintenance. Getting your oil changed every 3,000 miles is an example.
  • Predictive maintenance: Condition monitoring is utilized to predict when failure may occur. Maintenance is planned based on equipment condition. A vehicle dashboard that displays how many miles before you run out of fuel is an example.

Whether you’re focused on energy saving technology or simply maximizing the useful life of equipment, predictive maintenance has been shown to be superior time and again. Reactive maintenance cannot prevent catastrophic system failure, and even regular preventative maintenance can result in unnecessary work. Even if we were to ignore these factors, though, the return on investment speaks volumes.

Implementing a preventative maintenance plan over a reactive plan can help you save up to 18% on costs. By upgrading from preventative to predictive, though, an additional 8-12% is saved. When moving from reactive to predictive, an organization could see savings climb to 40%.

When moving from reactive to predictive, an organization could see savings climb to 40%.

When it comes to HVAC maintenance, this can equate to huge savings. In fact, studies have shown that adopting a predictive maintenance strategy delivers an ROI equal to 10 times the cost of its initial investment. Equipment breakdowns are also reduced by 70-75%. To put it simply, predictive maintenance is the unparalleled winner.

Predictive Maintenance Disadvantages

As with everything in life, there are pros and cons that come standard with the use of predictive maintenance. The most glaring disadvantage is the initial investment cost. Installing diagnostic equipment on HVAC or other systems and then training staff to implement predictive maintenance programs will of incur higher startup costs than just buying the system without any condition monitoring in place.

Of course, the improved return on investment typically makes the higher price acceptable. Unfortunately, hotels and other small businesses don’t always have the budget to cover these initial costs. Therefore, it’s preferable to find retrofit solutions that can accomplish the same feats as other solutions that require invasive installations for condition monitoring.

Another issue with predictive maintenance – especially for those trying to convince others to make the change – is that savings aren’t immediately seen by management. This maintenance strategy is a long-term approach that faces the same hurdles as other forms of sustainable energy practices. Management may not be “down in the trenches” to see these savings, but in this situation, the facts simply don’t lie.

Long-Term Advantages of Predictive Maintenance

The decreased costs and improved ROI with predictive maintenance is only part of the story. The most obvious long-term advantage is the increased operational life of the equipment. If you’re investing in green hotel technology, this advantage is enough to justify a predictive strategy. Your company will also recognize an increase in productivity up to 25% and a 35-45% reduction in costly downtime.

Your company will also recognize an increase in productivity up to 25% and a 35-45% reduction in costly downtime.

One of the more obvious long-term benefits of predictive maintenance – one that even doubtful managers or owners will recognize – is the ability to schedule maintenance in a manner that will reduce or eliminate overtime costs. Knowing exactly when maintenance is required also means you won’t need to order parts until they’re necessary – removing the possibility of inventory overload.

Latest Approaches to Predictive Maintenance

Reliability Centered Maintenance (RCM)

The RCM approach was developed with the base philosophy that “not all equipment is equal” in a facility. This is because in reality, different equipment will differ in terms of design, operation and importance to a facility’s safety and to its existing process.

Hence, the RCM approach maximises the use of limited resources by evaluating and prioritising predictive/preventative maintenance on more expensive and high-impact equipment/systems (e.g. chiller systems and their network of valves). At the same time, equipment that is less costly to replace and of low impact to your facility’s operational reliability will undergo reactive maintenance instead.

Typically, RCM will still involve a higher weightage of predictive (45-55%) and preventative (25-35%) maintenance on facility equipment with less than 10% of equipment on reactive maintenance plans. However, this allows your facility to better match existing resources to needs while maximising your operational reliability at lower costs.

Dissimilarity-Based Approach

In many facilities, predictive maintenance is used on a “homogenous cohort” or set of identical appliances (e.g all A/Cs in a building or all elevators in a building) by comparing the performance of each individual appliance against a standard behaviour. However, if you look at each appliance in isolation, you might be missing out on key patterns that only show up at the cohort level.

More specifically, a contemporary study in 2017 demonstrated that you can more accurately predict faults by finding inconsistencies in the pattern of differences across a homogenous cohort. This is based on the fact that a cohort of identical appliances (e.g. the A/C in the hallway VS the A/C in the guest room VS A/C in reception) should behave differently from each other in a consistent way.

Figure 1: The four coloured curves represent the behaviour over time of four appliance with respect to two sensors. In normal operation, one expects the four behaviours to be similar (if not equal), therefore the relative distances should show modest variations over time

Figure 2: Appliance 2 shows an anomalous behaviour, which is likely to be connected with an imminent fault. This is clearly reflected by a change in the relative distances.

Images taken from A dissimilarity-based approach to predictive maintenance with application to HVAC system

This approach has two important benefits over other standard approaches. First, you easily overcome the challenge of trying to identify the right set of factors/variables from the raw data that can most accurately predict faults in each appliance.

Secondly, by looking at concurrent mutual differences instead of absolute values of isolated appliances, there is no worry that the fault prediction model will be affected by seasonal trends or other external factors/biases.

Based on such promising results, it would be exciting to keep a look out for future studies on how the dissimilarity-based approach could also predict the remaining lifespan of equipment to further advance your predictive maintenance programs.

Summary

By now, you have a firm grasp on how predictive maintenance can reduce your overall costs and make every dollar you spend go further. If you’re on the fence about using technologies such as SensorFlow for your HVAC optimization system – or if you need the Cliff Notes version to take to your manager or finance director – keep in mind these facts:

  • – Over half of maintenance expenditures are lost to ill-advised reactive measures.
  • – Predictive maintenance can cut costs by 12% over preventative and 40% over reactive measures.
  • – Average ROI for predictive maintenance is 10 times your hotel’s initial investment.
  • – Costly equipment breakdowns are cut by 70- 75%, ensuring you get the most of the full expected life of your investment.
  • – Reduced man-hours required to monitor equipment thanks to automatic condition monitoring.
  • – Overall productivity increase of 25% and 35-45% reduction in downtime.

When it comes to saving energy, improving maintenance and increasing ROI at your hotel, SensorFlow is the obvious choice. Since it’s a retrofit solution, you’ll also avoid the high costs of implementation related to more traditional wired solutions.

Additionally, the ongoing cost savings will become apparent all while remaining eco-conscious with this green hotel technology. In a world where corporate responsibility matters, there’s not much more you could ask for in an HVAC optimization and automation solution.

References

Profit Margins in Hotel Industry (2018)

Operations & Maintenance Best Practices Guide (2013)

A dissimilarity-based approach to predictive maintenance with application to HVAC systems (2017)