Predictive Maintenance: Driving Reliability and Cost Savings in Industry 4.0

Predictive maintenance, which can identify maintenance issues in real-time, allows machine and vehicle owners to perform cost-effective maintenance and determine it ahead of time before the unit fails or gets damaged. When used correctly, predictive maintenance can extend the lifespan of industrial assets, reducing cost and increasing availability.

Spain has the second-largest high-speed rail network globally, with over 4,900km (3050 miles) of high-speed railway lines. And Renfe, the rail operator, promises on-time arrival on all its AVE (Alta Velocidad) trains across the country.

Until recently, to provide such service and availability, a rail operator needed to have a significant number of trains on standby to cover any unforeseen failures and maintenance issues.

However, Renfe is now able to keep over 99% of its high-speed trains operating at all times thanks to real-time monitoring, predictive maintenance, and on-demand component replacement.

Picture courtesy of Renfe.

Siemens, which builds and maintains most trains, uses a combination of thousands of sensors, edge computing, and real-time analytics to predict potential breakdowns and determine the best time for maintenance or component replacement. When a system or component requires maintenance, the system schedules the necessary downtime and makes sure that both the technicians and replacement parts are available to avoid any delays.

Because of this razor-edge technology, delays caused by technical failures with more than 10 minutes occur on average only every 1.5 million kilometers. The trains are available for operation 99.94% of the time.

Additionally, since components and systems are repaired or replaced only when needed, there are significant savings in parts and labor costs. Some elements can continue to be in service after their designed operation time.

Predictive maintenance helps industries realize the benefits of digitalization

Preventive maintenance is based on regular checks and replacing specific parts at scheduled intervals. While this has been effective in many industries and has helped avoid more expensive repairs, it creates a waste system. Many replaced parts are still in good enough condition to continue functioning correctly.

Unlike preventive maintenance, predictive maintenance replaces only the required part at the time necessary. Not only does it detect machine conditions that will lead to failure, but it also estimates the amount of time before that failure will occur, allowing service to be planned.

This concept is now starting to be applied to industrial assets. One of the most mentioned challenges to advance in Industry 4.0, is the large investment in legacy machinery. Now, many heavy machinery manufacturers are looking into a “rent” model, where the asset is leased to the operator, which pays for the real use of the machine. Maintenance and spare parts are part of the service.

Using real-time monitoring and predictive maintenance, the manufacturer can determine when it is necessary to send a technician to perform a maintenance task or instruct the machine operator to replace a component that is about to fail. This way, the asset’s downtime is reduced to the minimum and ensures the machine’s best performance for many years.

Predictive maintenance is coming to consumer products, including cars and appliances

Traditionally, a vehicle receives preventive service when it reaches a certain amount of driven distance or a specific amount of time since the last checkup, whatever comes first. This system is the one used by car manufacturers and dealers for decades.

Today, the price of most electric cars doesn’t include the battery. The battery is there but leased from the manufacturer. Instead of paying for fuel, many EV owners pay a monthly fee based on how much they use the car. Sensors installed on the battery, engine, and charging ports send the information to the manufacturer. When the battery is no longer able to hold a reasonable charge for the car user’s needs, it gets replaced.

Those replaced car batteries can have a second life, usually for ten years or more, as electricity storage for renewable energy or to balance the grid. After that, the units are taken apart, and the materials are reconditioned to be used again in new batteries or to manufacture other parts.

While there is a definitive move to electrification, most cars sold today are still using internal combustion. Car manufacturers and dealers are looking for ways to optimize maintenance by continuously monitoring all the vehicle’s critical systems. Using connected sensors and onboard analytics is possible to predict when maintenance is required.

For example, if driving in certain conditions requires continuous braking, using up pads, and brake fluid, the car subsystems alert the driver and schedule service way before those brakes fail.

The vehicle’s sensors can also measure environmental conditions, such as humidity, temperature, weight, and stress, plus other measurements such as pavement surface and tilt. The collection of all those data points can provide manufacturers and service centers with complete information about the environmental conditions of where the vehicle has been operating, something precious when evaluating advanced maintenance needs.

Appliance manufacturers are looking at the same concept. Unlike heavy machinery and other industrial assets, most home appliances are designed to last five to ten years. The reasoning behind it is that consumers want an inexpensive product and are used to buying a new one when it fails.

If washing machine vendors, for example, were able to receive a steady revenue by renting the units to consumers, they would make their products with better materials and components that lasted longer and fit them with an army of small sensors to provide continuous monitoring of the machine’s health.

When the sensors embedded in the machine detect that a critical component will fail soon, the unit will schedule a maintenance call, the necessary parts, and a technician to make the change. The user no longer needs to experience a broken appliance and have to wait for a repairer to arrive, diagnose the problem, and return later with the replacement parts.