Industry 4.0 Aspects – Machine Interference and virtual sensors

Posted on: April 20, 2015 by Wolfgang Thronicke

Industry 4.0 describes the concept of connecting machines, products, IT-services, and people in the production process and beyond. The realization of this concept targets an agile, highly reactive and efficient production even for small lot sizes with increased automation and "smartness". In this blog, I will explain how the "new" connectivity has additional benefits in the operation on the shop floor.

Machine connectivity in Industry 4.0 means that machines communicate a bunch of information related to their status and current activity and also receive supportive data and control instructions to perform their tasks efficiently. A rather interesting side effect of this is the machine interference. This interference is caused by physical proximity and actually is a crosstalk of signals which can be detected by the respective other machine.

In the past this effect was most unwelcome since each production machine has been regarded as singleton without any notion of its environment. So every measurement taken from a machine was subject to filtering this unwanted “noise” in order to properly operate the machine itself and control it.

Now we are approaching the notion of connected machines and an intelligent operation control on the shop-floor. So there is now a lot more data available from the machines as well as the capacity to process it in more sophisticated ways.

What has been “noise” can now be used to extract the causing signal which can be assigned based on the proximity information to the machine generating this signal and this is exactly the point where it gets interesting: With this property there is now a virtual sensor for a machine nearby available and this allows to exploit this new sensor with regard to all available information in a smart industry setting.

Consider the example of being able to measure the vibration signal of your target machine. It is quite simple (with the right algorithms and data) to detect when there are anomalies based on the current production step the machine is performing. Let us consider two scenarios how this can be used:

In the first scenario, there may be a pre-Industry 4.0 production machine, which has limited sensors and thus allows “only” the traditional ljmited control and feedback. On the shop-floor this machine is now enabled to participate through its new virtual sensors in a more state-of-the-art operation of monitoring and anomaly detection. This reduces “blind spots” in a smart factory setting which are to be expected when a gradual transformation from the traditional shop-floor takes place and not everything is updated and replaced at once (what every management of a running production would not do anyway because of the costs involved).

In the second scenario these virtual sensors can be used to detect tampering of machines on the shop-floor. Since the coupling is physical a compromised machine may itself claim to be working properly but utterly been driven to a destructive setting but the actual mode of operation can be detected using the virtual sensors. Basically this allows to check the plausibility of the feedback a machine is giving, where a cyber-attack may be a worst-case scenario, but even the detection of faulty sensors using this type of crosstalk-driven control would allow quick and effective maintenance to minimize downtime and the risk to produce faulty products.

The vibration signal in the example is of course only one possibility to create virtual sensors. By combining different physical values like temperature, sound and energy consumption a complete new layer of virtual sensors on the shop-floor augment monitoring and control functions leading to a smart operational control in the company.

This aspect shows, that the benefits of Industry 4.0 are not only direct but also create a lot of additional potential to improve the production area. And while Industry 4.0 is closely related to the Internet of Things concept, this "pattern" of exploiting the interference between devices can also be exploited in such applications.

Share this blog article

  • Share on Linked In

About Wolfgang Thronicke
Chief Technology Officer and member of the Scientific Community
Wolfgang Thronicke is CTO, group and project leader in national and European projects in the German ATOS innovation centre C-LAB. Since 2012 he is member of the Atos Scientific Community. With a 12 year background he is expert for software engineering, project definition and management for public funded and commerical projects. His current work topics include industry 4.0, cloud, IoT, and AI.

Follow or contact Wolfgang