While most companies recognize that, because of increasing competition, they need to look into modern technologies to support their production at every level, many companies also struggle to define a clear path from their current state to the desired outcome.
It doesn’t help that terms such as “predictive maintenance” and “prescriptive maintenance” allow for multiple interpretations and therewith different expectations from stakeholders.
To allow for a common understanding and therefore support decision making, we will use this blog to focus on prescriptive maintenance, break it down into 3 different levels and provide clear examples and use cases for all.
The first level of prescriptive maintenance relates to the use of ready-made and relatively simple scenarios of recommended actions. These actions are taken based on the occurrence of specific conditions, such as availability of spare parts, personnel or backup machines. These scenarios are developed on the basis of expert knowledge and “non-dynamic” data (for example the construction of a machine) and is most often performed at the machine or node level.
This first level does not require the use of advanced AI algorithms.
Example 1: In the event of detecting an impending failure of an electric motor, the prescriptive system estimates the time to failure, its criticality level, checks the competence base of maintenance employees, their availability, as well as the availability of spare parts, and, consequently, will inform all actors of the process to speed up the repair process.
Example 2: If an impending failure of a pump that is operating in a redundant structure is detected, the prescriptive system will recommend the user to switch to a spare pump and repair it.
Example 3: To service a piece of equipment with an overheating bearing, you could either take for instance 20 minutes to grease the bearing, and the result from the fix would probably last for two days. Or you could replace the bearing, which would take three hours but last for two years.
The second level of prescription, like the first one, comes down to the use of ready-made and relatively simple scenarios of recommended actions. However, in this case they are taken on the basis of complex data analysis. These types of scenarios are developed on the basis of expert knowledge, “non-dynamic” data and “dynamic” off-line data such as vibrations, temperature, pressure.
Level II prescriptions are most often performed at the level of the production process and do require the use of advanced AI algorithms.
Example 4: If an impending boiler failure is detected, the system assigns its data layout to one of the previously identified systems that are associated with a specific failure type and defines further preventive actions.
Example 5: If an impending turbine failure is detected, the system recognizes the characteristics of its vibration, on this basis identifies their cause, and selects one of the ready scenarios for further turbine control, the purpose of which is to stabilize its operation.
Example 6: A pump manufacturer will recommend specific operating design conditions such as discharge pressure and temperature, but there is a lot of variability in process operating conditions and also in the composition of the fluids. Prescriptive analytics can consider these conditions and make recommendations accordingly.
The third level of prescriptive maintenance includes the use of dynamic, complex scenarios of recommended actions, which are taken on the basis of complex data analysis. These scenarios are developed on the basis of all available data, in particular online production and business data.
Level III prescriptions are most often performed at the level of the production process and also require the use of advanced AI algorithms.
Example 7: If an impending pump failure is detected, the system will check the time to the next scheduled downtime and recommend that the pump speed is lowered by X, which will allow it to run continuously until the planned downtime. In addition, the system will calculate the production losses due to the reduced pump speed and compare them with the losses due to immediate downtime.
There are multiple drivers for companies to adopt a prescriptive maintenance approach.
Automation: As more automation is used within manufacturing processes, the speed of response that is required to address maintenance issues is going to get faster.
Economics: It’s getting more and more complex to make the right decision from an economical perspective. Knowing what can fail or when it might fail simply isn’t sufficient. It is necessary to have enough information to fully understand the options for maintenance, as well as the financial implications of each option.
Workforce changes: Experienced maintenance personnel and process operators are retiring, and their successors expect their tools to be smart and assistive.
Operating conditions: Assets do not only fail by their own means, but also because of how they are operated.
Asset performance: A higher level of sophistication is required in the way asset and process data is organized. The traditional plant historian and analysis tools have not been adequate to ensure asset performance. IIoT and analytics platforms are unique in their ability to ingest years of operational data and massive quantities of unconventional data scattered through different systems of record.
When implemented correctly, Prescriptive Maintenance solutions offer clear and tangible business benefits to not only the Maintenance department, but also Plant Management, and QHSE:
To make it simple. In the rest of the text, I will use the common abbreviation of Predictive Maintenance, which is PdM.Usually, if you don’t know what’s going on, it’s about money. Okay, it’s a gross simplification of a complex topic, but the conclusion is correct. There are also many emotions along the way, because […]
Attend our Live Webinars and learn all about how the latest AI-based technologies optimise your industrial processes, increase uptime and reduce your maintenance costs