By using operational data from industrial robots and implementing maintenance processes at the strategic level, equipment managers can achieve so-called predictive maintenance—linking maintenance techniques with real-time information from different devices and machines to perform maintenance on demand. This not only reduces downtime and increases output but also eliminates time and resources wasted on unnecessary maintenance.
By implementing predictive maintenance—rather than reactive maintenance—costs can be reduced throughout the equipment's lifecycle, potentially allowing most production facilities to significantly improve their profitability. This helps optimize energy use, reduce equipment downtime, and achieve improvements in other areas.
Why is predictive maintenance needed?
For production facilities with aging or even obsolete industrial robot equipment, maintenance procedures often lead to unnecessary costs, such as downtime, energy waste, and labor costs. Traditional maintenance procedures, which involve regular routine maintenance, mean that operators may be performing maintenance on equipment that doesn't require it, resulting in a waste of time and resources; or replacing equipment that is still usable. With traditional maintenance procedures, if equipment is not maintained as required, even if there are warning signs of an impending accident, they may be ignored.
On the other hand, production facilities that already perform predictive maintenance on industrial robots based on actual needs will have a different frequency compared to scheduled maintenance. By utilizing data generated by infrastructure such as networks and interconnected devices to process factors such as energy efficiency, temperature, and output, operators and plant managers can determine which industrial robots are functioning normally and which may be prone to failure. Based on this, operators and plant managers can make decisions on when to perform maintenance, schedule equipment offline, or, under current conditions, keep certain equipment running continuously.
When industrial robots cannot operate at full capacity, but their output remains within the normal range of variation, factory production facility managers can utilize predictive maintenance to avoid actual downtime. For example, a battery production line might produce batteries at an astonishing speed, even exceeding the discernible range for the human eye. A 10-15% fluctuation in output from three machines is within the normal production range. However, by utilizing other monitored data, such as energy consumption, operating time, and temperature, operators can increase machine output by 10%, thereby saving significant costs.
Big data is the foundation of predictive maintenance
Networks, interconnected devices, and the data collected, monitored, and analyzed (often referred to as big data) form the foundation of predictive maintenance processes. This data infrastructure, along with data-driven intelligent information, is what we're currently discussing as the Internet of Things (IoT). The IoT encompasses physical objects and connected facilities that incorporate embedded technologies to enable communication, sensing, or interaction with internal states or the external environment, allowing for the monitoring of equipment throughout the factory. Factory managers, operators, or original equipment manufacturer (OEM) maintenance personnel for industrial robots can use the data and information provided by the IoT to switch the factory to a predetermined predictive maintenance mode.
IoTLink provides comprehensive solutions and products from sensors to big data platforms, enabling IoT for devices and providing data support for predictive maintenance at all levels, including data sensing, network communication, data aggregation, data storage, data analysis, and data application.
Predictive maintenance can leverage many types of data, including equipment uptime, temperature, energy usage, output, and much more, to improve decision-making and operations. For example, in a consumer goods factory, a machine might run continuously, maintaining stable tissue production, but its energy consumption could spike significantly before it fails. By monitoring the machine's energy consumption data, operators can intervene promptly when a spike is detected, preventing downtime. Scheduled maintenance, on the other hand, requires taking the machine offline, resulting in unplanned downtime during the product lifecycle. By utilizing current data related to machine operation, as well as historical data from past failures, operators can mitigate the adverse impact on factory operations.
Key steps in implementing predictive maintenance
Achieving predictive maintenance cannot be accomplished overnight; it requires a multi-layered and gradual approach. Below are three key steps to begin implementing predictive maintenance within a production facility. For specific plans and steps, please consult the data application experts at Xiamen Wutong Bolian:
■(1) Changing Procurement Priorities: To do a good job, one must first have the right tools. To leverage big data and the Internet of Things for predictive maintenance, one must have equipment capable of generating this operational data. While connected devices are becoming the norm, the procurement process must shift its priority from traditional equipment to connected machines capable of network communication. This shift may present challenges for organizations, as traditional equipment without network capabilities has an upfront cost advantage over connected, smart devices. Utilizing data generated by connected devices can mitigate losses from single-failure events and resulting production line downtime, thus partially offsetting the additional costs of procuring network-enabled equipment. Procurement decisions must be based on the overall lifecycle cost of use, not just the initial investment.
■(2) Utilizing Data Experts: Once the equipment is network-connected and has the ability to measure and monitor data, the production operations manager can collaborate with data experts to ensure that the equipment collects and uses data in the most optimal way. Data experts can improve data operations by evaluating on-site or even virtual scenarios. Data collected by networked equipment can be stored in the cloud and monitored virtually through a server-based model. When data is virtually stored, it can be accessed, analyzed, and controlled with the help and guidance of data experts to direct and implement predictive maintenance. This virtualization, as a service provided by data experts, can accelerate the implementation of predictive maintenance within the factory.
■(3) Pushing the right data to the right people: A key aspect of using data to drive information and achieve predictive maintenance is pushing data throughout the organizational structure to exert maximum influence on the decision-making process. Data must be stored at a specific organizational level, but it must be pushed down to the factory floor level for individual machine operators to use. Just as with push notifications and data via smartphones, production operations managers must consider making data clear and understandable when working to ensure that data is transmitted throughout the organization and pushed to operators on the factory floor from various channels.
For example, in industries such as coal mining, mining, and metals, weather conditions are a key factor in enabling predictive maintenance. If data acquisition facilities and systems for optimizing data distribution are in place, operations managers can notify on-site staff and operators when severe weather is imminent, without needing to assign dedicated personnel to monitor weather forecasts. Intelligent data infrastructure can show which equipment will suffer the most degradation due to severe weather, the current status of the equipment, and specific maintenance tasks that should be performed before the weather condition arrives. In any industrial sector, operations managers should ensure that data reaches the lowest level, such as the shop floor, so that relevant personnel can respond accordingly. This often doesn't require calling maintenance experts to the plant, but rather ensuring that every equipment operator can use this data to perform predictive maintenance and optimize performance.
Comprehensive predictive maintenance programs can bring significant benefits to plant operations. Plant and facility managers who effectively utilize predictive maintenance can gain substantial operational benefits and a competitive advantage. Once equipment is interconnected, all stakeholders throughout the plant must trust the conclusions drawn from this data to maximize the benefits of data-driven predictive maintenance, even if these conclusions challenge previous understandings of optimal production parameters.
