Reliability Curve, Understand the Behavior of a Machine

15/04/2020• Reading time: 1 min• Updated 30 days ago

Among suppliers and resellers, equipment breaks down all the time, and this generates exorbitant maintenance costs and even terminations of millionaire contracts. Most of the time, these failures happen as a result of time , where the probability of their occurrence will differ between the stages of the life cycle, whether its a piece of machinery and equipment or an operation. The curve that demonstrates this probability of failure over time is known as the bathtub curve (or failure rate curve).

Usually, an equipment has, at the beginning of its useful life, a high failure rate, due to manufacturing problems, improper installation, defective components, incorrect assembly. Over time, these faults are corrected, and the equipment enters a level of stability. Breaks, when they occur, are random. After a certain period, according to the conditions of use and aggressiveness of the environment in which the equipment is located, the failure rates begin to increase, due to the wear and tear of the components.

How does the graph work?

The identification of the failure rate curve of a component or an asset, helps to control the maintenance schedule, especially controlling the health of the equipment, warranty time and reliability in the choices of necessary measurements for the increase in the availability of the systems under management.

Analyzing the reliability curve, we can identify 3 distinct points: Rate of failure decreasing for infant mortality; constant failure rate for service life; and increasing failure rate without limit on wear.

  • Infant mortality: During this phase, failures occur due to some manufacturing problems, installation defects, design errors, incorrect assembly and inadequate components. We have a high failure rate at the beginning of the equipment’s operation.
  • Useful life: Over time, these faults are corrected and the assets enter a level of stability, with a stable error rate and, when they occur, the faults are random. In this period, the number of failures is less than in the infant mortality stage.
  • Wear period: According to the conditions of use and deterioration of the environment in which it is found, the equipment starts to show a considerable increase in the proportion of errors. This happens due to the wear of the components.

We can draw several reflections on this, for example, the fact that the equipment has a phase of infant mortality clarifies something that at first glance could seem paradoxical: Preventive maintenance can increase the rate of equipment failures.

When equipment is still in adulthood, an intervention carried out at a time of smooth functioning, even if it prolongs the life of the equipment, exposes the asset to a peak of higher failure rates, due to infant mortality.

Therefore, estimating the most suitable time to perform preventive maintenance is difficult, unless the condition of the equipment is monitored. However, this already characterizes predictive maintenance, and no longer preventive maintenance.

3 methods to reduce failure during the infant mortality period

Three methods are suggested to reduce failures during the infant mortality period:

1. Debugging

Due to the high failure rate in the initial period, debugging tests are widely accepted as an approach to detect failures before leaving the factory until the product population reaches a low failure rate. Defective equipment is discarded or minimally repaired, if possible. Ideal for suppliers to obtain concise feedback on their equipment before it reaches customers with an anomaly or problem.

A major problem associated with testing is deciding exactly how long and at what level of assembly the tools should be tested, for this there are technologies and sensors that collect data from the equipment in real time, providing maintenance teams with a complete overview of the health of these active within minutes.

2. Acceptance and reliability

Acceptance tests can be periodic assessments of the reliability of the production material, especially when any tools, parts or other characteristics undergo any alteration or change. Generally, while there are misleading risks of accelerated testing, the benefits of such testing can be critical.

3. Health and quality control

Quality control is related to the identification and processing of information of machine variables to avoid the occurrence of serious failures. For example, vibration analysis can be used to detect possible problems when a process is getting out of control and take the corresponding actions before failure occurs.

Technology to your advantage

Of course, monitoring the useful lives of assets and their behavior over time, must be done by a complete management tool. There are many softwares that guarantee the manager a wide and real-time view of the maintenance situation.

With a technological solution specialized in asset monitoring, it is possible to analyze the cost of maintenance and interventions, for future analysis of the costs obtained. It is also feasible to control the scheduling of work orders to organize maintenance, distribute the workload and enhance the team’s performance.

In this scenario, HSaaS (Hardware Software a Service) has emerged as a practical model for the implementation of more effective forms of maintenance, because it provides in a single service the link between what needs to be measured in the physical world and the experience that the digital world provides.

Understand hardware as the sensor to collect equipment data in real time, and software as a way to process this information and provide maintenance teams with daily feedback on the health and availability of your assets.

Having information to understand the bathtub curve and monitor your assets is one of the benefits that technology can bring. Check out other befits on our website, or request a demo below.


Icon - Compartilhe no facebookIcon - Compartilhe no LinkedinIcon - Compartilhe no Whatsapp

About the author:

Foto do Autor

Gabriel Lameirinhas

Founder and Co-CEO of TRACTIAN. Computer Engineer from University of Sao Paulo, Specialist in predictive and passionate about industrial maintenance.

Linkedin do Autor

Related Posts


Are you being surprised by unexpected breakdowns of your machines?

Enter your e-mail and find out how to reduce this cost with our expert!