HomeTECHNOLOGYIoT: Instructions For Using Production Data In Connected Factories

IoT: Instructions For Using Production Data In Connected Factories

What parameters and how could we power intelligent applications to improve a productive activity? The importance of knowing which measurement is representative of the process we want to study is linked to selecting the most efficient method to transfer it to information systems. Extending the Internet to things has been one of the points industrial research has worked on for years. 

Data collection is based on sensors or the generation of logic values ​​by control units (in the industrial field, we think of PLCs or numerical controls). At the same time, various protocols are available for transmission and sharing. Reference architectures (frameworks). More and more often, we realize how much data is a fundamental part of our daily activities:

  1. We monitor the oxygenation of the blood from the smartwatch.
  2. We track our packages from the phone.
  3. We can turn on the house’s heating knowing in advance that it will be cold.

What if we could do similar things with production data? What parameters and how could we power intelligent applications to improve a productive activity? To answer these questions, we must start with the awareness that data today represents a precious resource in connected factories.

The Push Of Industry 4.0

The advent of the various national plans “Industry 4.0”, then “Enterprise 4.0”, and now “Transition 4.0” have pushed manufacturing companies towards the digitalization of production processes. All subsequent regulations to date have included the interconnection of assets as a fundamental requirement, underlining the need to use documented protocols and standards to facilitate data integration with third-party systems.

Third parties or systems are not necessarily related to the production context in which machine tools, and more generally the production technologies, operate. Although the sum of manufacturing integration had often remained anchored only to the concept of CAD / CAM, recent technological developments have opened Pandora’s box of IT technologies (Information Technologies) applied to production processes.

The companies, therefore, found themselves studded with new terms, which, although present also in previous years, now acquired the excellent thrust of the Industry 4.0 train. Data analytics, predictive maintenance, energy monitoring, MES, ERP, teleservice, adaptivity, digital twin, cybersecurity, and we could go on with other application crashes. Unfortunately, the risk that is run by many companies that still struggle to adopt these new tools is that of falling behind and losing the train of innovation. Let’s try to clarify how to navigate between these applications, identifying which and how much data it makes sense to recover—starting from the Internet of Things.

Sensors, Connection Protocols, Data Management Systems

Extending the Internet to things has been one of the points industrial research has worked on for years. And we don’t think of applications practical only to scientists in white coats; for example, let’s think about how today we control our sporting activities by monitoring routes and performances. Or let’s try to remember the last time we looked at the car registration document to understand how long it was until the next service, instead of waiting for the automatic message from the control unit. All these applications are based on the concept of the Internet of Things (IoT), or the idea of ​​extending connectivity to everyday objects, places and processes. 

This is a set of entities capable of hearing, transmitting and sharing data connected on public or private networks. The detection of these data is based on sensors or the generation of logic values ​​by control units (in the industrial field, we think of PLCs or numerical controls). At the same time, various protocols are available for transmission and sharing. And reference architectures (frameworks). There are, therefore, three leading players in this transformation:

  1. Sensors, for the detection of quantities from the field;
  2. Connection protocols, which instead define transmission rules;
  3. Systems for the management and consumption of data.

Each of these items has a separate chapter in the Industry 4.0 dictionary. Still, for a first approach, it will suffice to know that more than anything else, what will lead us to choose a type of sensor, rather than a specific protocol or a management platform, must be the expected result at the end of all the elaborations. It must be clear at the design stage what the ultimate purpose of the analysis we are going to do is because the data has no use if it is not consumed. Going up the food chain of data, transforming first into information, then into knowledge, requires introducing the last element: the cyber-physical system (CPS).

Smart And Connected Factories

In the industrial field, the Internet of Things concept is further strengthened and takes the name of Industrial Internet of Things (IIoT). The hinge is always the connection between devices, but we consider machinery, production orders, tools and everything that has to do with the production environment. A practical application of this concept can be found in monitoring systems, which exploit the data exposed by the equipment to analyze and reformat them into more digestible and immediate metrics, exposed through dashboards or control panels. 

Other uses may include planning input, resource management and planning support. Here is the definition in industrial terms: the data made available through the IIoT is consumed by a final application. This will generate a new interaction with the process, such as activating a specific subsystem of the machine tool to compensate for a process variation (adaptation to drifts). Or it could elaborate a metric of plant efficiency comparing the final quality of a product with the hours of production versus the hours of the possible output (OEE). 

And we could even go so far as to imagine a factory system that self-regulates its production according to the number of raw materials and consumables available in our warehouses. We could go on for a long time with examples of this type, but we will limit ourselves to condensing the importance of this structure in creating new applications and service levels.

Connected Factories, The Centrality Of The Data

But are these new applications? Not. The dream of a connected factory was already in the restless sleep of the technicians who first approached the idea of ​​a computer for production machines. Over the years, we have found various ways to refer to this concept. Without going too far, think that already in 1988 in Erba, there were interoperable production plants between different technologies, able to exploit artificial intelligence to optimize production. So what has changed since then? In short:

  1. Computing power has risen by orders of magnitude since the years in which these concepts were formulated. The numerical controls have such high speeds as to allow almost instant reactions in the production process, thus enabling closed-loop controls even for the most complex procedures.
  2. The data we can access is more robust, thanks to the development of resilient and shared protocols. Production requests go through e-mails and management databases, and we can more easily track and record the control variables in industrial processes.
  3. Writing an application requires a few hours of effort before a tutorial, and completely free and open-source programming environments are available, to quickly become programmers able to develop software for the industry.

Suppose today the availability of hardware seems obvious and the possibility of writing “Hello World” on any numerical control. In that case, it is not so easy to disentangle the various protocols and frameworks available to make these solutions efficient and timely. There remains a technical difficulty in selecting data, a savoir-faire of those figures who can correctly mix IT skills with knowledge of industrial processes.

So let’s go back to the starting point: the centrality of the data. Knowing which measurement is representative of the process we want to study is linked to selecting the most efficient method to transfer it to our IT systems. The road to connected and intelligent industries is now accessible to all navigators with a compass, wherever they point.

Also Read: More Than Process Mining Plus Business Intelligence

RELATED ARTICLES

Latest Articles