While conventional preventive machine maintenance is very statistics related (e.g. based on machine hours), unreliable and usually performed according to schedule whether it is needed or not. New technical solutions enable an early assessment of potential failures.

According to known Analysis:

  • 40% of current preventive maintenance costs are spent without effect on uptime
  • 30% of maintenance activities are carried out regularly without need 
  • 45% of all maintenance is not effective enough

Predicting specific failures can improve unplanned downtime by 70% and prevent loss of productivity and profit. Using real time Data to determine when exactly to perform a service has a major influence on profit, productivity and finally return on invest.

Industrial Business leaders should address the topic of predictive maintenance within their strategy and define the implementation process. But where to start?


Here are 10 steps to implement predictive IoT maintenance solutions within your industrial environment:

1: Get comfortable with the production processes

In production lines certain machines might be the bottleneck, causing the majority of downtime due to random problems. Gathering detailed information on reasons for a downtime, enables you to specifically apply a predictive maintenance solution to the bottleneck machine or part.   

2: Understand the value of existing data

Current machinery have embedded technology to control the machine and also gather data from these machines. This data is often collected but not analyzed and is not used for maintenance planning. Access to the already collected data by an application or an integration into an existing ERP System can enable predictive maintenance without separate hardware installations.  

3: Start Small

Implementing predictive maintenance solutions does not mean a huge investment. By identifying and eliminating bottleneck elements solutions can be specifically targeted. A maximum return for the smallest investment.

4: Execute a Proof of Concept

A Proof of concept allows to demonstrate the viability of the solution, to confirm that the integration is effective and proves of the potential future value of the solution.

10 Steps to implement predictive maintenance in your industrial enviroment 2

 

5: Start minimally invasive

Predictive maintenance IoT solutions can be implemented without interference on the machine control systems. Sensor technologies enabling vibration analysis, thermography or ultrasound can be installed on the external structure, decreasing any risk of security breaches.

6: Keep Security in mind

The level of security for the  solution must always be assessed depending on its criticality and on sensitive information away from competitors. The complexity of technical solutions such as hardware encryption of the data depends from case to case.

7: Set evaluation criteria for continued success

What criteria have to be achieved in a certain time frame in order to get your return on investment?  Is it regular failures on the main production line that should be reduced? Is it the ability to order key parts in advance and avoiding downtime entirely? Clear objectives ensure your ability to measure the success and return on invest.

8: Get the right stakeholders involved

The most likely cause of failure is not having the key stakeholders fully engaged with the strategy. From the management to the machine operators, all participants need to be involved in the strategy and know how it will affect the organisation. A predictive maintenance solution should influence the overall efficiency, addressing the topic and fear of job automation is critical for the acceptance within the team.  

9: Adress Cultural Change  

Industrial manufacturing is often driven from a culture established decades ago. Changing this culture takes time. A new culture is when everyone in the company understands the power of data. A data driven culture allows companies to capitalize on the power of technologies.  Companies with a new digital culture can become increasingly data-driven in their decisions.

10: Act

Once the teams have the data they need to deal with the upcoming maintenance, they have to act based on the information provided. It’s vital to put a work order defined by the data into the system and execute it in time. Software applications can support the execution of the operative process.