Learning Center

Information for building effective maintenance programs

Predictive vs Condition-Based Maintenance

Predictive and condition-based maintenance (PdM / CbM) are both maintenance types that occur before breakdowns happen. As such, they are both forms of proactive maintenance and are designed to increase reliability and decrease downtime.

The primary difference between them is the way in which maintenance is measured. Predictive maintenance relies on precise formulas in addition to sensor measurements (temperature, vibration, noise), and maintenance work is performed based on the analysis of these parameters. In this way, predictive maintenance is a very exact form of maintenance because it predicts future maintenance events.

On the other hand, condition-based maintenance relies only on real-time sensor measurements. Once a parameter reaches an unacceptable level, maintenance workers are dispatched. This means that condition-based maintenance systems perform work only in the moment it is needed.

Both predictive and condition-based maintenance can be expensive to initiate, though they both justify their upfront cost by saving money on downtime and equipment maintenance. Condition-based maintenance in particular can be expensive due to the cost of maintaining sensor devices, so it’s best used on critically important equipment.

Differences between predictive maintenance and condition-based maintenance

Preventive Maintenance Condition-based Maintenance
Definition Predictive maintenance (PdM) is work that is scheduled in the future based on analysis of sensor measurements and formulas. Condition-based maintenance (CbM) is work that is performed at the exact moment when measured parameters reach unacceptable levels.
Trigger Predicted date Measured parameter level
Cost Medium/High Medium/High (startup cost)
Cost Savings 25% to 30% Dependent on amount of equipment using CbM
Resources Needed
  • Maintenance software for scheduling
  • Maintenance scheduler (for larger organizations)
  • Condition monitoring software
  • Condition monitoring tools and sensors
  • PdM training
  • Maintenance software for dispatching technicians and creating work orders
  • Condition monitoring software
  • Condition monitoring tools and sensors
  • CbM training
  • Maintenance need is predicted in advance
  • Reduces maximum amount of downtime
  • Improved automation of maintenance tasks
  • Maintenance work is performed only as needed
  • Fewer unplanned downtime events
  • Improved prioritization of maintenance time
  • Expensive to implement and maintain
  • Time-intensive
  • High cost of installation, training, and maintenance
  • Difficult to choose proper sensor equipment
Use Case An organization has assets with slow-speed bearings that frequently fail. Preventive maintenance is already in place but the organization suspects that assets are being over-greased. To perform maintenance with more precision, they use ultrasound analysis (good for slow-speed bearings). Now, work orders for greasing are only scheduled when certain ultrasound measurements are reached. An organization needs to make sure that a critical piece of safety equipment can be maintained while it is running at load. One critical equipment parameter is the amount of vibration produced. The organization decides to implement CbM so that maintenance is performed only when vibrations begin to reach unsafe levels. This way, the safety equipment can be run constantly during maintenance and only receives maintenance when necessary.
Key Takeaways

Predictive maintenance (PdM) is performed when needed based on measurements and calculations; condition-based maintenance (CbM) is performed at the exact moment a parameter reaches an unsatisfactory level

Both types of maintenance use condition monitoring and measurements

PdM can result in cost-savings of 25% to 30%, while CbM can save costs dependent on how much equipment runs on the system

PdM and CbM are both somewhat expensive, though PdM is much more precise

More Articles
Learning Center Topics