Centric connect.engage.succeed

Cyber-physical systems, complexity and emergence

Geschreven door Ben van Lier - 25 november 2014

Ben van Lier
In July 2014 a 40-tonne truck took its first autonomous ride in Germany. This autonomous truck is supposed to be able to navigate through European traffic independently by 2025. Thanks to all sorts of technological possibilities, the traditional physical truck is slowly but surely becoming a cyber-physical system.

One of these new technologies is the truck's special LED lighting. When the lighting turns orange, other drivers know that the truck is driving autonomously. If the lighting is blue the driver is controlling the vehicle. The truck is also equipped with radar sensors and camera technology to support the autonomous operating system (Highway Pilot). The cameras detect single and two lane roads and identify still or moving objects and subjects such as pedestrians. Although the current version is only a prototype that still needs to undergo continuous testing, it has already been designed to move at speeds of up to 80 km/hour [1]. This autonomous truck is an excellent example of a cyber-physical system.

Cyber-physical Systems

Cyber-physical systems (CPS) are traditional physical objects that have been equipped with internal (embedded) software and network interfaces. Cars, airplanes, TVs, but also power plants and refrigerators can be connected in networks such as the internet. With the help of its software and its ability to make connections, a CPS can generate data about its current status and context. The CPS can then give meaning to the data and exchange both data and meaning with its environment. After receiving and analysing the information, other CPS's in the network can then apply their own, new interpretation to this information and take action accordingly.

Through a continuous cycle of communication, analysis and interpretation and the resulting actions, cyber-physical systems can learn to interact with each other in networks. A report by Loughborough University [2] states that in this development, cyber-physical systems form 'a natural consequence of the increased connectedness and autonomy of real-time embedded systems' (2013:30). By combining it with sensors, cameras and autonomous operating systems, the traditional 40-tonne truck is being changed into a cyber-physical system.


An individual CPS - such as a truck - is a complex entity, comprised of several components that communicate with each other (engine, navigation system, controls, radar sensors, camera’s, etc.). Each of these components is also a system in and of itself. In order to make decisions about which route to take, the autonomous truck must continuously exchange data with various internally and externally connected systems. Moreover, the resulting combinations of data must be given new significance each time. Hence, new information and the meaning given to it within the system continuously lead to new decisions (to be taken by the system).

This creates an autonomously functioning unit. A self-driving vehicle that continuously makes traffic decisions based on communication and interaction between its components. And all this without endangering passengers and other road users. A CPS in the form of an autonomously driving truck can therefore be viewed as a complicated entity. As Cilliers  (1998) states: 'Some systems have a very large number of components and perform sophisticated tasks, but in a way that can be analysed accurately'. (1998:3)


When individual cyber-physical systems (autonomous trucks) communicate independently in a larger system (the traffic system) and interact with other systems (people, other vehicles and trucks, sensors in roads, GPS systems, traffic lights), a system of systems is automatically created. In such a scenario, the individual system communicates not only with its own components, but also with technical and social systems in its environment. This gives rise to a so-called socio-technical system, 'an engineered system which includes a combination of technical and human or natural elements' (2013:9).

Such a system of systems can be seen as a complex entity. According to Cilliers [3], complexity arises 'as the result of a rich interaction of simple elements that only respond to the limited information each of them are presented with. When we look at the behavior of complex systems as a whole, our focus shifts from the individual element in the system to the complex structure of the system. The complexity emerges as a result of the patterns of interaction between these elements.' (1998:5) The properties of the whole can no longer be attributed to the individual autonomous vehicle but develop naturally within the whole of (autonomous) vehicles that communicate with social and other technical systems in their environment.

In the coming years, such complex systems of systems - also called ultra large scale systems of systems - will rapidly grow into a scale that can hardly be imagined (e.g. Internet of Things, Advanced Manufacturing). Ultra large scale (software) systems (such as large quantities of autonomously functioning trucks) that mutually interact and communicate with other systems in their environment give rise to new questions regarding dimensions, such as 'lines of codes, amounts of data stored, accessed, manipulated, and refined, number of connections and interdependencies, number of hardware elements, number of computational elements, number of system purposes, number of routine processes, interactions, and emergent behavior, number of overlapping policy domains and enforceable mechanisms and number of stakeholders.' (2013:12)

Emergent properties

The report published by Loughborough University points out that such large-scale systems of systems are characterised by properties of the whole that come into being automatically and can no longer be attributed to the individual components. The autonomous development of such properties is one of the biggest issues regarding complex systems. What choices will the mutually connected trucks make when something goes wrong? How do we prevent unexpected or undesired connections in such an enormous quantity of collaborating algorithms? Or, as John Holland [4] wrote: “The Hallmark of emergence is this sense of much coming from little.' (1999:2).

We need to learn more about the complexity that comes about when new technologies are connected with people in networks and communicate and interact with each other. But above all we have to learn whether and how we can handle properties that arise automatically in these new systems. We need to learn how to deal with unexpected and unanticipated properties of this large entity determined by technology. Learning how to handle uncertainty stemming from technological development requires a new vision of society that we are creating with the help of technology.

Ben van Lier works at Centric as an account director and, in that function, is involved in research and analysis of developments in the areas of overlap between organisation and technology within the various market segments.

[1] http://www.dailymail.co.uk/sciencetech/article-2766635/From-reclining-seats-aircraft-style-radars-Mercedes-takes-cover-self-driving-TRUCK-set-hit-roads-2025.html
[2] T-Area-SoS. SoA report, Work Package 2, deliverable D2.1, Revised Version. Trans-Atlantic Research and Education Agenda in Systems of Systems. Loughborough University, 2013
[3] Cilliers P., (1998) Complexity & Postmodernism. Understanding complex systems. Routledge, Oxon, Canada. ISBN 0-415-15287-9
[4] Holland J.H. (1999) Emergence. From Chaos to Order. Perseus book group. ISBN 978-0-7382-0142-9

Holland J.H. (1999) Emergence. From Chaos to Order. Perseus book group. ISBN 978-0-7382-0142-9

    Schrijf een reactie
    • Captcha image
    • Verzenden