Optimization is not the only algorithmic tool that supply chain companies need to solve their planning and decision-making problems.More May 2, 2019
The power of predictive maintenance
Authors: Özgün Aydın from ICRON and Lieneke van Boxel from CQM
For most of us, maintenance is an invisible process that we never notice or think about – until one of the machines that we rely on has an unexpected breakdown. Whether it’s an airplane or train or a piece of manufacturing equipment, we always assume that the machines in our lives will continuously be able to run smoothly and successfully – until the moment comes when they can’t.
Indeed, we usually fail to plan for breakdowns. That is why sudden machine failures that result in service disruptions – the airplane or train delay that throws our travel schedule off or the manufacturing machine malfunction that causes late delivery of goods to customers – are such a source of frustration. But what if there was a way to anticipate and eliminate these breakdowns before they actually occur? There is, and it’s called predictive maintenance.
Comparing the main maintenance strategies
Broadly speaking, the purpose of all maintenance – of course – is to prevent failures as much as possible. Traditionally, companies have practiced two basic approaches to maintenance: preventive and corrective. The former strategy seeks to proactively prevent machine failures from occurring by performing planned maintenance tasks at predetermined intervals. Corrective maintenance, in contrast, is a reactive strategy that aims to remedy a malfunction that has already occurred as quickly as possible.
Both approaches have their shortcomings. Corrective maintenance does not actually help to prevent machine breakdowns and downtime – as it happens after the fact. Although preventive maintenance does reduce the chances of a piece of machinery failing by conducting maintenance according to a prescribed schedule, it typically leads to high maintenance costs (for labor, spare parts, etc.) This is due to the fact that with preventive maintenance, the maintenance tasks – in many instances – are not absolutely necessary, but are routinely performed at specific times as a part of a preset plan.
Predictive maintenance (also known as condition-based maintenance) utilizes data and advanced analytics algorithms to continuously monitor the actual condition of a company’s assets, predict how long these assets can continue to function properly and when they will need maintenance, and recommend the best time to schedule and perform maintenance activities.
With predictive maintenance, companies can foresee and avert asset breakdowns and downtime while also avoiding unnecessary preventive maintenance (and its associated costs) –thereby improving both productivity and profitability.
Utilizing predictive maintenance technologies
Unfortunately, manual techniques and tools like Excel are not up to the task of conducting predictive maintenance-based planning and operations – to do this, an automated, algorithm-based planning and optimization platform (with a robust analytics engine) is required.
Such a platform is capable of compiling and processing all the historical and real-time data (from IoT devices, back-office systems, and other sources) on a company’s assets, and utilizing that data along with advanced analytics and machine learning algorithms to automatically generate maintenance plans and schedules that:
- Take into account the past performance (including duration and frequency of routine maintenance and unplanned downtimes and types of maintenance required) as well as the current condition of a company’s assets.
- Forecast the frequency and duration of the assets’ future downtimes and the maintenance tasks that will need to be performed during these downtimes.
- Predict and provide the optimal time to conduct specific maintenance tasks – grouping together various maintenance tasks that can be performed simultaneously (thus minimizing asset downtime and maximizing maintenance efficiency).
- Ensure the availability and optimize the utilization of maintenance resources such as specialized engineers, equipment, and spare parts.
- Integrate each asset’s maintenance activities into the company’s daily operational or production schedule to optimize each asset’s ongoing utilization.
- Can be dynamically and optimally revised if an unexpected breakdown occurs and corrective maintenance is required – to reduce the impact of disruptions on regular operations.
By enabling companies to make optimized plans and decisions about which maintenance tasks to do and when to do them, an algorithm-based planning and optimization platform helps predict and prevent asset breakdowns, decrease maintenance costs, and improve asset efficiency – and, ultimately, the company’s bottom-line performance and levels of customer satisfaction.
Realizing the benefits of predictive maintenance
Predictive maintenance technologies give companies the power to accurately forecast asset failures and downtimes and perform preventive maintenance only when it is absolutely necessary – thereby decreasing both asset downtime and maintenance costs.
With the predictive maintenance technologies, companies can transform themselves by taking a more proactive (and profitable) approach to maintaining their assets – rather than a reactive (and inefficient) approach where they are constantly, frantically scrambling to respond to sudden, unexpected breakdowns.
Operationally and financially, predictive maintenance technologies can bring significant benefits by enabling companies to keep their assets – such as planes, trains, or manufacturing machines – rolling along smoothly without interruption, keep their maintenance costs down, and keep their customers happy.