Factories are bringing intelligence to the production line in a number of ways. Most are focused on integration between machines on the factory floor and leveraging sensor data for actionable insights.
Texas-based software firm Epicor identifies certain examples: shop-floor scorecard systems for reviewing performance, touchscreen human-machine interfaces (HMIs) for reporting human-related tracking data, and automated ‘Poka-yoke’ systems for preventing errors.
Andrew Robling, the company’s senior product manager, also runs through seven enabling technologies (systems, protocols, and techniques) that enterprises are using integrate intelligence into their production processes. They are as follows.
1 | Machine protocols
OPCA and MTConnect are industrial protocols for the exchange of data between shop floor equipment and software applications used for monitoring and data analysis. “For connecting to the PLC on machines to capture key details, like machine state, cycles, parts produced and scrap,” explains Robling. As such, they are essential components in the process to glean data from machines and apply data insights to them.
2 | Sensor devices
Say no more; in many cases, additional sensors must be retrofitted to industrial machines to take performance / process measurements, which can be fed into edge / cloud analytics models to bring new control and intelligence operations. Robling explains their purpose, “to track and maintain temp, pressure, vibration etc; enabling alerts for errors that fall outside of an acceptable range”.
3 | Visual systems
This covers everything from analytics dashboards, which make insights digestible and actionable, to more sophisticated visual interfaces, including AR/VR systems, as tools to interact with machinery and processes, and ‘digital twins’, which can render manufacturing functions in 2D and 3D, and reflect modifications in the manufacturing process across their entire lifecycle (PLM), so changes are accommodated in design, engineering, production, and utilisation. Robling puts it simply: “Spatial modelling (digital twins) to confirm products / parts are made to spec.”
4 | Human-machine interfaces
A variation on the above; visual rendering of processed data, in dashboards and other interfaces, enable factory workers to automate and control factory machines in new ways, and make good on the insights the AI functions deliver. These “human machine interfaces (HMI) connect people and machines, devices,” says Robling.
5 | Execution systems
The heart of the factory; the whole discipline of data analytics in manufacturing is geared towards the manufacturing execution system (MES), which forms the control plane for production, covering its dynamic automation and monitoring. Robling calls them “MES solutions for full statistical process control and quality control.”
6 | Edge analytics
“Triggering next-best-actions to act on errors, such as automatically rejecting a part that doesn’t meet specifications,” comments Robling. The point is, often, with fast-paced machine processes, the need for speed is everything. Cloud-based analytics, process machine data in far-off data centres via even broadband connections, does not cut it. The intelligence must be brought to the edge.
7 | Intelligence systems
Business Intelligence (BI) is an umbrella term for the applications and disciplines for collecting, integration, analysing, and presenting enterprise data. In manufacturing, BI systems provide historical and predictive views of process and machine performance, as well as live views operations. They handle both structured and unstructured data, as required, to bring focus to operations, and the rise of AI has burdened them with greater traffic and responsibility.
This article is taken from a new report from Enterprise IoT Insights, called ‘AI on the line — how advanced analytics and artificial intelligence are transforming the production line’. The report is available to download for free, here.
Table of Contents
