Cyber-physical systems unite the physical world of machines with the virtual world of the Internet. They are the core of Industry 4.0.
The concept of cyber-physical systems arose from the idea that a workpiece can be fully simulated from the design phase through to production and autonomously identifies its own optimal production process. This means production can be completely automated; the only human activity required is in the design phase and in defining the production conditions and parameters. Dr Heinz Derenbach, the Executive Vice President Technical and Engineering in the Corporate Sector for Information Systems and Services of Bosch GmbH said: “In an extreme vision, this unfinished material already knows for which customer it is intended and carries with it all the information about where and when it will be processed. Once the material is in the machine, the material itself records any deviations from the standard process, determines when it’s ‘done’, and knows how to get to its customer.”
Autonomous thanks to embedded electronics
This is made possible by microelectronic systems possessing their own computing capability, communication and networking chips as well as the necessary sensors and actuators. They are thus able to perform all data acquisition, processing and output functions independently. These embedded systems are integrated into larger systems or objects, such as a workpiece as described above. In contrast to mechatronic systems – that is to say, systems in which mechanical, electronic and IT elements interact, as have been in operation for a number of years already in the automation sector – the sensors and actuators used in cyber-physical systems are “smart” solutions. In addition to their actual main functionality, they incorporate smart functions such as signal processing or control, as well as Internet-capable network communication ports such as Ethernet.
Utilising information from a wide range of sources
That ability to use the Internet, and so access a variety of software services – all based on open global standards – is the key new feature of cyber-physical systems. Because it then becomes possible to utilise information from a wide variety of different applications, organisations and fields. Cyber-physical systems are interconnected among themselves, as well as being connected to higher-level and lower-level systems. As a result, production is no longer managed by a central instance, but through a large number of small, distributed units. The units are becoming continually more intelligent based on the integration of powerful processors. The aim is that they should be able to make independent decisions within the workflows.
Autonomous decision-making
Behcet Acikmese, Assistant Professor at the University of Texas in Austin, is working with a multi-disciplinary team of psychologists and computer scientists to extend the capabilities of cyber-physical systems to incorporate decision-making. To do so, he is studying human behaviour: “Although we are far from autonomous decision-making that can handle many things that humans can do effortlessly, we are making progress,” Acikmese said. “Humans have efficient ways to observe the operational arena and make decisions that are robust to address potentially dynamic uncertainties. We would like to analyse human decision-making in relevant mission contexts and use insights obtained to design robust autonomy protocols.” His wider research goal is to develop software for embedded systems which is able to make decisions in real time.
Development methods and tools
Notwithstanding that progress, the development of complex cyber-physical systems for applications which impose high demands in terms of reliability is very complicated and costly based on present-day development techniques. Prof. Michael Butler, of the University of Southampton and Scientific Coordinator of the Advance consortium, said: “It is widely recognised that development costs will become prohibitive for future systems unless significant improvements are made in the methods and tools used for systems engineering. Its aim was to develop methods and tools specially for the modelling of cyber-physical systems. The result presented in late 2014 was a system incorporating a development process and a freely available collection of tools. It enables errors in specifications or in the design to be detected at the earliest possible stage of the development process. The Advance Toolset has been launched as an upgrade to the existing open-source platform. The new features are helping engineers to cut costs in system development by delivering exact models for the simulation of system behavior.
(picture credits: Shutterstock)