Asset maintenance is a core operational function of any business that relies heavily on equipment and infrastructure, as such its been around in one form or another since the industrial revolution.

Over the years many different types of industrial maintenance methods have evolved, as companies try to find the sweet spot that keeps their operations running smoothly.

In this article we compare the different types of maintenance and why we at VROC believe AI powered Predictive Maintenance is the best method as we move into this new decade.

  1. Run to Failure (Breakdown Maintenance)
  2. Preventative (Scheduled Maintenance)
  3. Predictive Maintenance (PdM)


This is a reactive method of breakdown maintenance, that when applied to low-cost, non-critical equipment can be a sufficient means of operation. Run to failure can result in more downtime and higher maintenance costs due to the necessity to require urgent parts, however unlike preventative maintenance, equipment is left to run and is not disrupted unnecessarily for maintenance.

When moving forward into a Predictive Maintenance method, some run to failure may continue to exist for non-critical low cost assets.


Preventative maintenance is scheduled maintenance regardless of the equipment’s condition and output. This is a proactive method that can be either time or trigger based. The purpose behind a preventative approach is to prevent failures and extend the assets lifespan.

Whilst a preventative method sounds ideal, it can result in unnecessary costs, including higher labor costs and a costly spare parts inventory. And there is the question of the necessity of maintenance on equipment that is functioning and outputting sufficiently.

One of VROC’s clients recently changed from a preventative maintenance method to a predictive maintenance program, which resulted in a 50% reduction in their spares inventory along with a savings of 15-20% in maintenance costs (see our case study).


The predictive maintenance method looks to forecast and predict when failure is going to happen, and allow sufficient time for maintenance to be scheduled and parts to be ordered before the failure takes place.

This method monitors the condition of assets using IOT sensors which provide real time data. Coupled with historical data, baselines are created. When certain conditions are met on the equipment the team can schedule the maintenance in advance.

The VROC Platform uses AI and Machine learning technologies to learn from historical & live data, creating new models which help to predict when maintenance is required, but also the cause of the failure, the time to failure and how to optimize the equipment. When connecting an entire enterprise to VROC, the platform will identify non-linear causes of failure which may not have directly correlated when manually analysed. The vast amounts of data can easily be handled in next to no time at all.

Manual predictive maintenance can be performed by a reliability team of engineers and data scientists, however as we have learned from many of our clients this work can consume thousands of man-hours due the volume of data and the complexity which can cause some failures (see our case study).


Changing maintenance methods can be a daunting task as it effects the core of business operations and traditional processes. At VROC, we believe with the right team this digital transformation can be simpler than expected. Change management is critical, especially alleviating the un-founded fear that AI will replace jobs within an organization, when in-fact it allows staff to focus on critical components of their roles and can reduce the mundane elements.

Another stumbling block that can hinder a companies transition is the condition of their data. With big data can come big messes, and it is data accuracy that leads to accurate predictions. Thankfully data lakes exist that can house an array of raw data both structured and unstructured, and once set up predictive analytics can easily commence. Learn more about data lakes here.

Lastly there is still a perception that artificial intelligence is expensive. Most people don’t realize that the technology has been commercialized now for many years and is being applied in some form or another in a large number of platforms today, often without our knowledge. In VROC’s experience the cost output is often quickly realized by the benefits of production increases, and the reduction in inventory and maintenance costs.

In a world where production output is critical to the sustainability of many businesses, asset reliability teams need to be equipped with the most advanced tools and scientifically proven maintenance methods. Intelligent Asset Management Systems like VROC exist to equip reliability teams with insights that can help them optimize equipment settings and avoid failures whilst improving the equipment’s lifespan.

Ready to start your transition to predictive maintenance? Schedule your demo of the VROC platform here.

Useful Resources

You might be interested in

5 Data & AI Essentials for Australian Councils and Cities in 2024

Five Data and AI essentials that every council or city C Suite member needs to understand, plus helpful next steps you can implement straight away.

Read Article

Unlocking the Power of Time Series Data

Unlocking the power of time series data - How advanced analytics and AI are revolutionizing forecasting and prediction in industrial applications.

Read Article