The emergence of industry 4.0 in telcos is still diverse, depending on a company’s core business and departments. MNO network teams tend to focus on active equipment, applying vendor strategies that ignore infrastructure maintenance. Therefore, O&M teams are often left empty-handed with only the ability to correct issues and, occasionally, the capability to prevent them. The problem here with preventive maintenance is the manual scheduling is often based on what sounds “reasonable” (i.e., once a year, before or after winter) over what sounds “cost-effective”.On the other hand, towercos live with the knowledge that they need to keep their infrastructures in good shape for the lowest cost. To reach that objective, they have started to invest in predictive maintenance, where decisions are taken* based on data (i.e., date, equipment aging, running hours) not on a gut feeling. 

In its last study, Analysys Mason anticipated that predictive maintenance would be adopted by 86% of tower operators by 2023. The study reveals that predictive maintenance is one of the five key drivers for towercos to invest in data management systems.

At ITD, we see it as an opportunity to create more value for ClickOnSite users. We are convinced that predictive maintenance is not only a nice-to-have feature. It is a great addition to our clients’ competitiveness providing more comfort to their staff and, therefore, greater employee satisfaction.

Michael Blot, General Manager APAC at ITD, provides us with his experience on how to implement predictive maintenance.

* Note that ClickOnSite enables automated decisions that are crucial in its ROI. Events are not only predicted but they are scheduled automatically (with no mistakes & no-cost planning).

Implementing Predictive Maintenance

Predictive maintenance can bring large benefits to any organization (i.e., MNO, towerco) provided that its raw materials (data) are already reliable and organized.

Therefore, our first step is always to look at available repositories. Most of the time, we find spreadsheets and the issues (i.e., mistakes, redundancies) that go with them.

How does it work? We often start by setting up a clear data model (more and more, our standard TMF inspired data model).

Then, we put in place the necessary means to harvest data (i.e., manual entry, import, mobile data collection) seamlessly whenever it is possible (i.e., interfaces, RPA).

After we succeed in collecting data, it is time to process them through algorithms that are easily configured in ClickOnSite with no programming.

From that moment, ClickOnSite can generate predictions that are traditionally used in reporting or notifications for making decisions.

This is where ClickOnSite goes even further! Yes, it monitors the state of equipment. It predicts failure based on patterns. But, it also proactively fixes problems to optimize service management. It makes decisions for you! It creates work orders, suggests planning, and helps you learn from the past.

 

Leading by example

In the summer of 2020, the largest towercos in Myanmar launched an RFP to support them with predictive maintenance. Their goal was to carry out site maintenance based on field data for energy management (monitoring of generators & visit planning), fuel management (refueling planning), and environment management (site maintenance).

At first glance, ITD did not look like a favorite. We do not claim that we are the #1 solution (although, in our case, it might be true). Unlike other companies, we are completely transparent in the way we work. We do not use AI to describe basic algorithms. Instead, we use our practical understanding of what makes a technical department operate and our corporate wheel to create value for users.

So, at the end of the day, the client compared our technical and commercial offers. To our surprise, its staff spent a lot of time checking on prototypes and the capacity of contestants to deal with their use cases. They agreed that we were right on target! They saw what makes our offering different and disruptive. They understood the power of low-code and BPMN that enable “almost” unlimited configurations. And, they were convinced that we speak the telecom language.

A couple of months later, they are starting to enjoy ClickOnSite capacity to predict fuel consumption and establish refueling plans.

So, how does it work? When technicians visit a site (i.e., for maintenance, refueling, etc.), they capture the GE running hours on the ClickOnSite mobile app. Then, the app synchronizes with the ClickOnSite Servers and data is uploaded and processed automatically to predict the optimal date for the next refueling.

The running hours, as well as the GE aging, are used to predict but also trigger specific maintenance packages automatically.

 

Benefits of Predictive Maintenance

If poorly managed, maintenance can have an important impact on a tower’s profitability, SLAs, customer satisfaction, and lead to a significant revenue loss. Tower owners should not wait for that day to happen…

So, now is the time to adopt predictive maintenance and enjoy its numerous benefits:

Conclusion

Nowadays, it is clear all the facets of maintenance (“passive” i.e., towers / “active” or “functional” i.e., refueling) can benefit from the power of predictions.

It is proven to work on existing shelf solutions, so take time to compare and ask for prototypes. If we cannot predict your decision, we can foresee a shortlist of objectives at the end of your RFI process :-).

If you would like to know more, and need support designing your predictive maintenance strategy, do not wait to ask for a free demo!