Today, organizations with high-value assets, such as fleet management companies, are turning to predictive maintenance to control the ever-rising costs of maintaining their equipment.
What is Predictive Maintenance?
Predictive maintenance tools monitor the condition and performance of in-service equipment, such as a fleet of vehicles, to reduce the likelihood of failures. The aim of predictive maintenance is to help companies save money by reducing the frequency of maintenance tasks and minimizing the disruption to the business.
These predictive maintenance tools work by continuously extracting and integrating data from a variety of sources, including vehicles and their environments, analyzing historical and real-time data, and comparing and cross-referencing the various types of data. In addition, predictive maintenance makes use of machine learning techniques to predict hidden patterns within the data.
How Predictive Maintenance Works
To take advantage of predictive maintenance, a fleet management company must have a computerized maintenance management system (CMMS), or some other way to track the data from its vehicles, the right Internet of Things (IoT) sensors to feed the data to the CMMS, and engineers to keep the system running.
Using the data gathered by the numerous sensors installed on trucks and other fleet vehicles, fleet managers can predict imminent failures. This decreases downtime and reduces costs as planned maintenance is typically cheaper than maintenance done in an emergency.
Predictive maintenance systems for fleets look at the condition of the vehicles, the engines, and/or other components, enabling technicians to make repairs or replace old parts when necessary. Remotely identifying potential problems before they occur means less time diagnosing, and more time fixing.
For example, a predictive maintenance system can identify that a part in one truck from a large fleet is X days away from failing. Once this part is identified, a member of the data team sends a follow-up report to the maintenance team, detailing the best options for the time to replace the part as well as the service location.
Predictive maintenance is an interesting area of business technology since it combines IoT and big data, said Seth Lippincott, an analyst Nucleus Research in Boston, Massachusetts.
“First, users need IoT signals from their assets and second these signals need to be analyzed to determine what the likelihood of failure is for whatever parts or assets are being monitored,” he said. “The data cannot be analyzed through conventional analysis techniques and require methods that look for patterns and meaningful signals amidst the noise.”
Many fleet management companies already have their assets connected and are collecting performance data. By analyzing the historical data – rather than relying on guesswork –users are able to extract when failure is most likely, Lippincott said.
“As a rudimentary example: most engine motor oil recommends that it be changed every 3,000 miles a vehicle is driven,” he said. “By leveraging big data analytics focused on predictive maintenance, fleet managers could determine that the optimal time to change the oil is actually every 5,000 miles, reducing the amount of scheduled maintenance that’s required.”
And vendors are looking to enable customers by delivering services that provide the algorithms and compute power to analyze this data, according to Lippincott.
“In terms of value delivery, fleet management companies can realize less downtime and unscheduled maintenance and can better schedule the procurement of replacement parts as well as technician routing/utilization,” he said. “For enterprises with numerous, highly distributed assets, such as fleet management companies, the benefits can be significant.”
Advantages of Predictive Maintenance
Predictive maintenance offers the following benefits:
- Reduces vehicle downtime – allowing fleet managers to ensure that their vehicles are on the road, not sitting in the shop.
- Decreases maintenance costs – by identifying potential issues before they arise fleet managers are able to have technicians repair the specific components, rather than the parts themselves, and stop further damage.
- Reduces spare parts inventory – managing their vehicles and being informed of issues as they arise, enables fleet managers to order parts and other items “just-in-time” rather than keep massive inventories on hand.
- Maximizes uptime – by reducing unexpected faults, predictive maintenance results in maximum uptime.
- Decreases unplanned maintenance – by monitoring the vehicles and the way the drivers operate them, fleet managers can replace/repair parts before the vehicles break down.
- Lowers labor costs – knowing when vehicles will need to be serviced reduces unplanned maintenance, which reduces overall labor costs.
- Increases safety – by continuously monitoring the health of their vehicles, fleet managers can prevent critical issues that could be disastrous.
- Manages faults – enables fleet managers and maintenance teams to stay on top of the health of their fleets by offering real-time information with expert insights and recommendations.
- Reduces fuel costs – when vehicles are maintained properly and run optimally, they use less fuel.
Uptime is a primary concern for fleet managers, said Dave McCarthy, senior director of products, Bsquare Corp., an IoT solution provider, technology distributor, and system integrator in Bellevue, Wash. Since they are responsible for the operation of capital-intensive, revenue-generating equipment, any strategy that can reduce unplanned downtime is worth considering.
“The sophistication of maintenance strategies has been maturing. Most fleets have a preventative maintenance schedule, which is typically interval-based,” he said. “This could be triggered by hours of operation or miles driven. In either case, it doesn’t reflect the actual health of the equipment. It’s just a rough guess based on historical knowledge.”
However, predictive maintenance is forward-looking and based on actual operational data. It allows fleet managers to forecast pending failure events so they can manage their businesses around it, according to McCarthy.
“For example, if you could identify the leading indicators to a failure, it is then possible to monitor for those conditions,” he said. “Once identified, it puts the fleet manager in a position to schedule needed maintenance ahead of time – before the failure ever occurs. This enables fleets to transition from reactive response to proactive planning.”