Research from market insights provider IoTAnalytics indicates that making edge computing systems “intelligent” through the integration of smart tools is a key driver of the technology’s continued growth. Edge analytics is a major enabler of intelligent edge solutions, expanding its use cases by supporting low-latency, high-volume data operations.
The need for more flexible and cost-effective operations has long been a top agenda item for manufacturing companies. A 2020 survey by industrial automation provider Yokogawa revealed that 48% of respondents considered productivity a key focus of their digitalization strategy, and 40% considered operational efficiency a primary objective. The pandemic has further increased the demand for automated processes and technologies to keep businesses running smoothly.
The lasting impact of the pandemic has accelerated the industry's digital transformation. Edge computing plays a key role in facilitating this acceleration, but making the edge intelligent is crucial to maintaining its value. Edge analytics is the process of collecting, analyzing, and processing data gathered from IIoT devices directly from the edge. By processing data as close to the data source as possible, edge analytics enables manufacturers to improve efficiency and accelerate innovation. But how?
Accessing machine data
Big data has laid the foundation for Industry 4.0, but accessing it correctly continues to challenge manufacturers. Factory floors contain numerous different machines, all collecting data that can provide valuable insights. Retrieving relevant data in the right format is the first hurdle for manufacturers hoping to fully leverage their edge capabilities. Choosing which data to process and use to trigger local operations and which to send to the cloud for storage, model training, and historical analysis is crucial.
However, edge analytics controls more than just the volume of data. It also helps harmonize data by converting disparate datasets into a common format for machine compatibility and comparison. Factory floors have multiple generations of equipment, all collecting data in different ways. Many different data sources—such as PLCs, DCSs, historical data, and databases—and many different protocols—Modbus, MQTT, OPC, Siemens, and ABB—need to be processed in different ways.
Processing large amounts of data at the edge can prevent cloud systems from being overwhelmed and significantly reduce associated costs. By avoiding expensive cloud services, processing and storing data solely in the cloud can reduce costs by up to 99%.
Simplify industrial processes
Solving the data access problem is the first benefit of edge analytics for manufacturers, but how to fully utilize the collected data will be the next challenge. Research by global market analyst Forrester estimates that 60% to 73% of all collected data is not used for analysis. However, utilizing data in real time can improve machine performance and streamline operational efficiency.
Edge analytics data gives manufacturers the opportunity to evaluate data as it is generated and respond by deploying operations directly to machines to improve their performance. For example, a machine's operating speed or the amount of material it distributes can be modified instantly based on data collected from the next machine on the factory floor.
Choosing to do this at the edge rather than in the cloud makes this application possible. Keeping data local facilitates valuable machine-to-machine (M2M) communication between different generations of devices running on different protocols, thus streamlining the manufacturing process. Furthermore, keeping data local alleviates a common industry concern regarding security and data policy—distributing processing and algorithms, rather than distributing data.
Improve enterprise management
The efficiency of the factory floor impacts every aspect of business operations—a slowdown in production or equipment failure can cause significant disruptions throughout the supply chain. Just as edge analytics can connect machines and processes without sending data to the cloud, it can also integrate data into Enterprise Resource Planning (ERP) systems. ERP systems are business process management software that manages a company's financial, supply chain, operations, manufacturing, and human resources activities in one place.
ERP systems are increasingly moving towards event-driven architecture (EDA), which uses information to connect business functions in real time in response to “events.” These can be anything from customer requests to sensor readings to inventory updates. When an event occurs, an event-based ERP system uses a set of rules to ensure that relevant data is sent to all business areas that might need it.
Modern event-driven edge analytics software can serve as a connection layer between the factory floor and ERP systems, sending relevant data to other business functions in real time. This allows data collected directly from the factory floor to be used across multiple business areas to improve quality control, meet growing product demands, and prevent disruptions caused by unexpected equipment downtime.
Edge analytics is a key technology for fully leveraging intelligent edge infrastructure. By facilitating real-time communication between machines, processes, and other business areas to achieve more efficient production output, edge analytics allows manufacturers to maximize the potential of machine data, improving efficiency not only on the factory floor but also throughout the company's operations.
