When it comes to maintaining machinery, most organisations currently use preventative or reactive maintenance.
Preventative maintenance involves inspecting and replacing parts which have reached a certain age, while reactive maintenance is fixing something which has broken.
Neither of these are particularly efficient. Repairing something which has already broken means the machines is out of action until repairs are finished, disrupting - or even completely halting - workflow. Obviously, machinery can break at any point – leading to the possibility of something going wrong at the worst possible time.
Preventative maintenance may seem to be a wiser strategy, but in reality it’s still guesswork. You might be replacing (and wasting) a durable component which would have lasted much longer with one which is less reliable and could be damaged more quickly.
Thankfully, a third, more efficient maintenance method is emerging. The widening uptake of the Internet of Things (IoT) has led to the possibility of ‘predictive maintenance’.
How does predictive maintenance work?
You’re probably more familiar with the use of IoT in the home, using apps to control lighting, heating, appliances and other connected technology. The principle is exactly the same in industry.
This connectivity not only enables businesses to control machines, it also creates a multitude of data points. This data can be analysed by businesses using the machines – and also the company who designed and developed it. This data allows engineers to make more informed decisions about when a machine requires maintenance.
By creating and uploading algorithms to devices based on previous performance and also projected usage, engineers can ascertain when machine or part is likely to fail. Additionally, by using machine learning, devices themselves can detect patterns in their readings when compared to existing data from other machines. The devices then flag when it transmits data which could highlight an issue is occurring.
Most smart technology also contains sensors which diagnose performance. Known as ‘condition monitoring’, this equipment uses many techniques, including vibration analysis and infrared imaging, to detect any changes in device usage. There are even some machines that are able to self-repair based on these readings.
With this combination being relayed to engineers in real-time, they can make smarter decisions on when to intervene, and predict the optimum time to replace parts or entire machines. This is truly ‘predictive’ maintenance.
Predictive maintenance still underused
Despite the benefits predictive maintenance provides, its use is not yet widespread across the industry. This is partly down to transformation coasts and integration difficulties.
A survey by global management consultancy Bain & Company found that companies are having difficulties integrating IoT powered machines with existing operational technology and IT systems.
Transitioning from old machines which do not produce much usage data to new IoT enabled devices can be costly, as well as disruptive. It may be easier for companies to transition slowly, with small pilots and incremental progress, rather than try to overhaul machinery and processes in one giant leap.
Despite this slow progress, Bain & Company estimates that the industrial IoT market will be worth more than $200 billion by 2021. This demonstrates that, when done correctly, predictive maintenance will bring huge efficiencies to businesses.
What does this mean for engineers?
The evolution of IoT and predictive maintenance will undoubtedly change the role of maintenance engineers. While the self-repair function used by some smart devices has created suggestions that maintenance engineers are no longer required, the true picture is far different.
Whereas traditionally the role has been focused on manual processes, going forward engineers will be freed from repetitive, manual jobs. Instead, maintenance managers and engineers will focus on understanding what a machine is saying and using that data to improve operations. They won’t be required to inspect devices periodically themselves, but will become more proficient in analysing data and identifying when machines are losing efficiency.
Increasing adoption of predictive maintenance will alter the skills engineers need, with a greater requirement to understand the technology they are maintaining and proficiency in data analysis software. As with most uses of automation and AI, lower value, repetitive tasks will be done by machines, with engineers able to fulfil higher value roles.
If you are looking to recruit engineers to transform your organisation, or are an engineer looking for your next role, get in touch with your local Reed office.
