Blog Internet of Things

What is IoT-based Predictive Maintenance?

Predictive maintenance (PdM), also known as condition-based maintenance, is one of the hot topics when discussing the application areas of industrial IoT and smart factories. But what is IoTbased predictive maintenance and what are the benefits?

Maintenance has always been an intrinsic part of any production. A good maintenance plan ensures safe operations, equipment availability and reduces costs. Maintenance includes the troubleshooting, repairing, controlling and verifying physical equipment, and contributes to the improvement of industrial processes. According to Gouriveau et. al., the last 20 years have meant new requirements for maintenance in terms of quality, safety and costs. As industrial equipment become more complex, they requires greater competence in maintenance as well. [1]

Predictive maintenance techniques are used to determine when in-service equipment need repairment. PdM is based on real-time analysis of data from the industrial equipment [1]. The objective is to prevent unexpected failures without costly and inconvenient routine checks, or any unnecessary downtime.

Predictive maintenance requires condition monitoring. This means the continuous monitoring of machines during process, to ensure optimal use. Condition monitoring can be done continuously, periodically and remotely. First, production or machines are monitored continuously, with data collected on critical moments. Periodic monitoring gives analysis over the changing vibration behaviour. And finally, remote monitoring enables the equipment to be kept an eye on from a remote location. [2]

Data collection and processing is one of the main components of PdM. The data collected can be based on physical phenomena such as vibration, temperature, pressure, voltage, light dispersion or humidity. Other kind of data can be, e.g. process deviations, raw material quality, control settings, or machine specification. Technology has made it possible to monitor machines and equipment continuously. This is done by using different kinds of sensors to assess the degradation and predict the failures ahead of time. [3] In recent years, the the availability of wireless industrial internet of things (IIoT) devices has made the process and setup easier and faster.

Read how Haltian’s wireless IoT sensors are used to monitor connecting components in electrical substations here.


The benefits of predictive maintenance are vast, and differ from case to case, but here are some of them listed:

  • Reduction in the number of breakdowns
  • Increased reliability of production processes
  • Improvement of personnel safety
  • Reduction of periods of inactivity for the equipment
  • Incresed performance of the company
  • Increased equipment lifetime [1]
In addition to this, predictive maintenance functions can often be made when the equipment is on service. Thus minimizing any interruptions to operations.


Preventive maintenance (PM) means maintenance that is carried out before the failure occurres. Now, this might sound very similar to predictive maintenance, but PM techniques include a schedule for periodic checks. These checks are performed to all monitored elements, whether they actually need maintenance or not. According to Gouriveau et al., this “predetermined maintenance can lead to overcare, that is, an excess of useless interventions, and thus financial wastes for the company.” Predictive maintenance is more dynamic, as it takes into account the current condition and attempts to predict the equipment’s state evolution in time [1].
Another problem with preventive maintenance is the possibility of breakdowns between the routine checks. In these cases, unplanned maintenance (or run to failure maintenance) is performed when something is already broken which, of course, leads to unknown costs depending on the broken element.


In conclusion, it can be said that well planned and implemented PdM is the foundation of any maintenance program. Investing on predictive maintenance for critical operational functions could save you a lot of money in maintenance costs alone, not mention the increased safety. Though predictive maintenance won’t erase all routine checks, it can certainly reduce them.
Luckily, new technologies and digitalization has made the optimization of maintenance possible. The question now is how to get started? 

Get to know Haltian’s Smart Factory Solutions or read more about all our Commercial IoT.


[1] R. Gouriveau, K. Medjaher and N. Zerhouni, From Prognostics and Health Systems Management to Predictive Maintenance 1: Monitoring and Prognostics, 2016

[2] “Condition Monitoring of rotating machines,” [Online]. Available:

[3] N. Amruthnath and T. Gupta, “Fault Class Prediction in Unsupervised Learning using Model-Based Clustering Approach,” 2018.


Yrttipellontie 1D
90230 Oulu, Finland

Finland Capital area

Haltian Oy
Aalto University Campus
Metallimiehenkuja 10, 02150, Espoo


Switch board: +358 10 322 3400
Fax: +358 10 322 3499