September 8, 2015

Predictive maintenance - the new future growth engine

Anu Maajärvi-Kosamo

Lead system analyst, Tieto

In April 2015, the Prime Minister's Office published a report called ('Turning Finland into the Silicon Valley of the industrial internet', publication series April/2015 by Government’s analysis, assessment and research activities), which shed some light on the future prospects of the industrial internet. 

One particular story set in the near future stuck to mind:

You are brushing your teeth in the morning. While you are doing this, you make efficient use of your time and simultaneously read the latest news displayed on the digital screen on the bathroom wall. Your smart, networked toothbrush recognizes the status of your teeth while you are brushing and detects some issues with your oral health. This information is sent directly to your dental nurse, who, if needed, sends you an invitation to a dental appointment.

As your teeth are subject to continuous monitoring, any changes in their condition can be addressed right away. This reduces the need for larger dental care work in the future. Because the service you use reduces the total costs of dental care, social security subsidizes the costs of the service you use. Best of all, your pearly whites will stay with you longer.

A well-maintained machine will be more productive in the long term

Predictive maintenance is not a new invention. The better the machine is maintained, the longer its life expectancy. Skipping maintenance may save money in the short term, but, in the long term, a well-maintained machine is likely to be more productive. Timely maintenance and replacement of parts improve performance.

Machine or equipment downtime  is very expensive for an industry. Annoying as it is, systems and equipment are often not repaired until they no longer work.

Predictive service and maintenance aim to predict and prevent various breakdowns and service disruptions and to respond to them in good time. When we analyse real-time information collected from machines and equipment, we can respond to future incidents proactively. The mass of data can be analysed for causal relationships, which enables more accurate control of service and maintenance.

Towards smart spare part business

Some time ago, one of the leading Finnish manufacturing companies  came to us with a business challenge.  Our customer's business comprises the manufacture of industrial machines and their spare parts for different sectors of industry as well as various maintenance services. Our customer wanted to investigate the spare part wearing more thoroughly to be able to help own customers to optimize the usage -  and control the costs related to it. There was  a lot of data  available about the spare parts, but the challenge was to be able to utilize it for the customers' benefit?  How should that data be used  to  create competitive advantage and increase sales? What is needed to make the  spare parts smart so that it becomes easier for customers to manage their equipment base and understand and predict the wear and replacement cycle of the parts? How can damage and expensive production shutdowns caused by sudden breakage of machines be prevented?

Added customer value through  machine learning

Together with the customer our data scientists  grasped the challenge by introducing  new perspectives and innovative thinking  into the customer's own research.  The spare part sensor data could be analysed - and understood  better than before. This was done using machine learning. Finding correct algorithms and filtering methods made it possible to analyse the wear curve of a spare part and create more accurate forecasts. Any service interruptions could also be detected and eliminated before the actual analysis.

Our customer will implement the created solution in the near future.The analysis and modelling of data carried out in collaboration with the customer will offer  the customer a significant competitive advantage in the future. By collecting the data created automatically in spare part sensors and analysing it using predictive analytics methods, we were able to more accurately  predict factors affecting the wearing of the spare parts.

Potential benefits of industrial internet solutions are expected to be extensive in the near future. According to the report referred in the beginning, these  solutions can- when successfully applied -  help create a centre of competence in Finland – a bit like the Silicon Valley – and establish conditions for business growth and value production, which benefit society as a whole.

Smart, predictive dental care, as described at the beginning of this article, is not yet in place for us to enjoy. In manufacturing industry, however, similar solutions are being developed and introduced. Companies that are at the forefront of developing new solutions will benefit the most.  In the future, also taking care of people, will depend more on equipment planned to continuously monitor us. This development offers many opportunities since both people and machines need predictive maintenance.

Predictive maintenance

Categories:

Stay up-to-date

Get all the latest blogs sent you now!