Moving from reactive to predictive maintenance
The Internet of Things (IoT) has changed the manufacturing industry by refashioning business models, increasing overall output, and automating processes. Internally, the IoT wave has simplified plant equipment maintenance and externally it has enabled manufacturers to enhance customer support services.
Where reactive maintenance was once the norm, manufacturing companies can use predictive maintenance to proactively handle problems even before customers realize they exist. The end result is a hassle-free customer experience rooted in reduced downtime and maintenance costs. With IoT enabling manufacturers to raise their customer-centricity quotient, it comes as no surprise the global spending on IoT solutions is expected to increase from $29 billion in 2015 to $70 billion in 2020. The next wave of manufacturing—smart manufacturing—is fast approaching.
Responsive system enables proactive performance
Calendar-based and reactive maintenance practices have progressively given way to predictive maintenance regimes—more than a quarter of manufacturers are already tracking product performance through predictive maintenance applications. With manufacturers expressing a gradual interest in the IoT, the operational predictive maintenance market size is estimated to grow from $582.3 million in 2015 to $1.88 billion by 2020.
Manufacturers are collating masses of data from assets along the value chain and analyzing them to harness insights on asset utilization, failure prediction, and asset lifetime. For instance, whenever a specific component or part of an asset functions sub optimally, the asset sends the requisite information to a central system via machine-to-data technologies and remote monitoring applications.
Building a customer-oriented servicing ecosystem also requires manufacturers to focus on maintaining individual products in their installed base. Gathering installed base information in real time can vastly improve the value they generate for the customer, as well as boost after-sales operations’ profitability. The remote monitoring of installed base also creates a possibility of providing services remotely.
Most importantly, it generates a new source of revenue while boosting customer satisfaction—higher market penetration of such services enables manufacturers to counter-balance the revenue drop during economic downturn when products do not sell. This information enables asset management teams to detect possible points of failure or weaknesses before they result in sudden downtime or malfunctions—bridging together physical assets, people, and processes seamlessly.
Encouraged by these functionalities, many manufacturing companies are adopting IoT production and monitoring processes to lower costs, boost safety, manufacture robust products, and improve overall equipment effectiveness.
IoT production and monitoring processes can nevertheless offer other benefits in the longer run. Manufacturing plants can achieve far greater visibility into the manufacturing floor, increase productivity, and enhance profitability. Predictive maintenance can make it simpler to optimize uptime reliability and schedule downtime so customer orders can be met in a timely manner. Additionally, as predictive maintenance programs involve using a central database that keeps a tab on spare parts, asset health and customer orders, tracking costs and ensuring proper financial management will also become less of a burden.
Moving toward a model of efficiency
With the IoT emerging as a strong transformational force, it is becoming indispensable for manufacturing companies to choose the right predictive maintenance model for plant equipment maintenance and customer experience enhancement purposes. While the IoT storm is particularly boosting the use of composite analytical models, manufacturers can also opt for conventional analytical models to realize both these ends:
- Conventional analytical models distinguish between repairable and non-repairable parts to forecast failure based on period of use. These models leverage historical time-to-failure data to determine the future failure of parts. In situations where manufacturers do not have the privilege of detailed information from sensor data, these models prove particularly useful.
- With composite models, the difference is the factors that cause parts failure are also identified. This feature is available over and above the usual characteristics of the conventional analytical model.
For manufacturers, the question then is which model better suits their specific needs. Towards this, several issues warrant consideration. First, the availability of data is a crucial determinant of which model is appropriate. When data on events leading up to parts failure and information on parts failure itself is available, composite models are more suitable. On the other hand, when just claim data is retrievable, conventional models tend to be more suitable.
Others factor influencing analytical selection are the nature of parts—repairable or non-repairable get altogether different treatment for failure prediction—and the stage of the parts lifecycle (introduction, growth and maturity or decline).
With the right model, manufacturers can take their predictive maintenance efforts to new levels. As more players realize this, we will see smarter, profitable, and efficient manufacturing businesses populate the sector.
Ravishankar Kandallu, Tata Consultancy Services (TCS). This article originally appeared on the Industrial Internet Consortium’s (IIC) blog. The IIC is a CFE Media content partner. Edited by Chris Vavra, production editor, Control Engineering, CFE Media, email@example.com.