Predictive maintenance is not new, but today more than ever with advancements in the industrial Internet of things (IIoT) and artificial intelligence (AI), predictive maintenance can result in significant savings for manufacturers.
Manufacturers utilize predictive maintenance to minimize the possibility of downtime by using sensors to monitor operational conditions, storing historical data in the cloud, and performing analytics. This makes it possible to service equipment based on actual wear and tear instead of scheduled service visits. Think of it as bringing in your car based on real-time measurement of actual fluid levels, vibrations and belt thickness, instead of every few thousand miles.
For industrial equipment, more efficient maintenance translates into big savings. It’s a combination of longer equipment life, more efficient use of field technicians, and avoiding expensive downtime, which can lead to product delays and potential safety hazards. When predictive maintenance is in place, machines on the factory shop floor or appliances in a customer’s home can even evaluate their own performance as well as order their own replacement parts and a field technician when there are indications that equipment needs to be serviced. Predictive maintenance can even make use of algorithms based on big data to predict future equipment failures.
When administrative procedures related to ordering and installing new parts are triggered automatically, cost savings can also be experienced in the back office. For example, a machine could sense that a drill bit is wearing out and automatically order a new one, alert the technical service department to send a field service representative, and forward the purchase request for a new part to the ERP system. By automating manual, error prone, labor intensive administrative functions, manufacturers can experience an additional level of efficiency.
But one of the biggest stumbling blocks to predictive maintenance is making data flow smoothly from machines to ERP systems in order to achieve a high level of security and reliability with a low level of latency. However, these barriers are coming down one by one because manufacturers have a strong incentive to invest in predictive maintenance because of the strong payback.
Predictive maintenance ROI
Data from the US Department of Energy indicates that predictive maintenance is extremely cost effective. Putting a functional predictive maintenance program in place can yield remarkable results: a tenfold increase in ROI, 25%-30% reduction in maintenance costs, 70%-75% decrease of breakdowns and 35%-45% reduction in downtime. When savings are expressed per labor hour, predictive maintenance costs $9 hourly pay per annum while preventive maintenance costs $13 hourly pay per annum.
The reasons are simple. Reactive maintenance work costs four to five times as much as proactively replacing worn parts. When equipment fails because there is a lack of awareness of degraded performance there are immediate costs as a result of lost productivity, inventory backup, delays in completing the finished product, and more.
A study by The Wall Street Journal and Emerson reported that unplanned downtime, which is caused 42% of the time by equipment failure, amounts to an estimated $50 billion per year for industrial manufacturers. Even after production begins again, the costs of interrupting operations continue. According to the Customers‘ Voice: Predictive Maintenance in Manufacturing report by Frenus, approximately 50% of all large companies face quality issues after an unplanned shutdown.
In addition to savings, predictive maintenance can also result in competitive differentiation. When machine data can be used to perform predictive maintenance with a high level of precision, manufacturers can focus on differentiating products using digital capabilities like self-healing based on an awareness of technical health. A manufacturer’s value can be measured not only by the quality of its shop floor processes, but also by how it protects its assets. Predictive maintenance can be a selling tool to show customers the manufacturer’s built-in ability to extend equipment life and improve the efficiency of maintenance procedures.
Predictive Maintenance in Action
UPS claims it has already saved millions of dollars by implementing a predictive maintenance solution to reduce breakdowns and extend the equipment life for their fleet of trucks. Managing over 55,000 drivers and more than 100,000 vehicles globally, UPS has already stored over 16 petabytes of data including information about engine performance and the condition of the vehicle as well as speed, number of stops, mileage and miles per gallon.
Siemens has successfully implemented predictive maintenance for NASA’s cooling systems at Armstrong Flight Research Center situated on Edwards US Airforce base in California. The system monitors the performance of fans, pumps, air handlers, and cooling towers while gaining insights into potential reductions for maintenance and operating costs. Every time there is a significant status change for a piece of equipment, automatic notifications are sent to NASA and an analyst for review.
Deutsche Bahn (DB) and Siemens have launched a pilot application for the predictive servicing and maintenance of the high-speed Velaro D trains. Siemens utilized a special data analysis center, the Mobility Data Services Center in Munich, to predict potential equipment failures.
There are a number of vendors that advertise the ability of their components to initiate their own service calls including Cummins Power Generation who alerts homeowners and technicians automatically by mobile apps if their generators could experience any potential equipment problems or have specific service requirements.
The IT Challenge of Predictive Maintenance
There are several pieces of the puzzle that need to be put in place before there is a fully working predictive maintenance system. Machines, devices, sensors and people need to connect and communicate with one another seamlessly. There needs to be a virtual copy of the physical world in order to make sense of all the data to conceptualize the information. The most sophisticated solution technologies, such as AI, need to be deployed to support decision making and problem-solving, making cyber systems as autonomous as possible. Here are some specific requirements.
• Bringing your organization into the fourth industrial revolution – Industry 4.0. To implement Industry 4.0, you need to start with state-of-the-art sensors whose technologies can be linked to large amounts of data in almost real time. Sensors need to be able to monitor conditions with a high level of reliability in real time to provide meaningful data. Many sensors are limited by antiquated technology and need to be updated.
• A robust ERP system that can handle unstructured data is the key to unlocking the value of collected data. There is a need to integrate and aggregate data collected from separate solutions from different product manufacturers in order to create meaningful data to create insights. After aggregating the necessary data, you need a system that can store the data, process it and analyse it.
• For predictive maintenance solutions, this is not a small thing, for it’s easy to imagine that there are, for example, different delivery truck models in a fleet, various turbine makes and manufacturers in an electrical plant, or different heating and cooling manufacturers in an industrial park.
Hand-coding integration can limit the IT department’s ability to respond to business changes. This creates the danger that IT can be perceived as slowing down a company’s progress by becoming a bottleneck, using valuable resources for integration projects. Integration platforms can be a better alternative to hand-coded integration, because they provide a more flexible environment that can handle multiple integrations with different systems that frequently require updating. They are optimized to deal with different vendors’ technology stacks and to optimize between stacks. Fault-tolerance, resilience and elasticity as well as monitoring and performance management capabilities are built-in so that all data can be managed in a standard way across systems.
• An integrated ERPMES- platform empowers Industry 4.0 by connecting core systems to enable manufacturers to have all the necessary data combined to gain the needed insights to achieve higher levels of quality and productivity. Enterprise resource management (ERP) manages the business of manufacturing products, manufacturing execution system (MES) controls the production process itself.
• Integrating your ERP and MES systems can enable alerts for immediate actions to prevent downtime. Maintenance management is much more effective if brought to the operations level, where it can be integrated with production processes closer to where things actually happen.
• ERP and shop floor systems operate on different interpretations of “real time” — ERP focuses on strategic management of the business, while MES is associated with the punctual execution of the process.
• By integrating ERP and MES, manufacturers can sync customers, orders and inventory data with the shop floor in order to meet actual production requirements, and for reconciling material consumption for better planning.
• When managers can visualize, monitor and document manufacturing processes on the shop floor, including equipment conditions, they can detect changes in status and put the administrative functions in motion to rectify the situation, such as ordering new parts and scheduling a service call. The MES system can be used to populate the ERP system with the quantities actually manufactured and scrapped as well as performance levels. Manufacturers can then use the realtime business management information provided by the MES to fine-tune production schedules. When the systems work in unison, and not just in parallel, manufacturers can respond more quickly to fast changing operating conditions.
• According to the 2017 Market Guide for ProductCentric Cloud ERP Solutions by Gartner, “By 2020, at least 35% of new productcentric ERP deployments in large enterprises will be software as a service (SaaS) loosely integrated with on-premises manufacturing execution systems (MES).” This emphasizes the importance of shifting these two disparate systems from a mere coexisting mode to an integration mode.
• Integrating your logistics and MES systems. Logistics systems, including Warehouse Management Systems, Transportation Management Systems as well as ERP systems all need to be integrated with manufacturing systems to automate an endto-end business process for equipment maintenance including scheduling the closest available field technician with the necessary tools and expertise.
• Investing in this multisystem integration can have a quick ROI due to significant financial benefits. For example, one pharmaceutical company used an integration platform to trace lot numbers and chemical quantities during each step of the manufacturing process, accelerating the compliance with regulations. Integrating specialized equipment with back office equipment also reduced inventories due to better demand forecasting.
• Secure data is a must. Data needs to be collected and shared with a high level of security to protect companies’ intellectual property. Detailed manufacturing equipment information including production processes can be an important part of a company’s competitive advantage, and so can destinations for trucks for delivery companies .
• Customers’ financial and personal data needs to be protected, especially with regulations such as GDPR that enforces high penalties when customers’ data is used without their explicit permission.
• Built-in scalability. When all the sensors are monitoring equipment condition, fluid quantities, vibrations and heat, huge amounts of data need to be collected, stored, and analyzed. When a middleware or data management solution is used in this type of scenario, the solution requires builtin scalability as well as a high level of performance and reliability to collect, filter, process and share vast volumes of data.
Despite the technological stumbling blocks that we have briefly analyzed here, predictive maintenance is a vital part of maintenance management of the future. Manufacturers who succeed in tackling the integration issues and automating both manufacturing processes and maintenance can benefit from a massive financial advantage by taking their operations to a whole new level of efficiency. In time, manufacturers of dishwashers, clothes washers, and maybe even cars can sell hours of service, because of their high level of confidence in the operating efficiency of their equipment, taking away the risk of equipment failure away from the consumer. And all of this can only happen when predictive maintenance enables people and machines to communicate with a high level of security and efficiency.
• Manufacturers utilize predictive maintenance to minimize the possibility of downtime by using sensors to monitor operational conditions.
• One of the biggest stumbling blocks to predictive maintenance is making data flow smoothly from machines to ERP systems in order to achieve a high level of security and reliability with a low level of latency.
• In addition to savings, predictive maintenance can also result in competitive differentiation.
• Hand-coding integration can limit the IT department’s ability to respond to business changes.
• Integration platforms can be a better alternative to hand-coded integration because they provide a more flexible environment that can handle multiple integrations with different systems that frequently require updating.
To learn more about Predictive Maintainance in detail register your intrest in the Industrial IoT European Summit, taking place in Munich, Germany or The Industrial Internet Innovation Summit taking place in Shanghai, China.
This article was originally sourced from InfoQ