Cyber physical systems for Industry 4.0

This is an older blog post from 2018. Read our updated 2019 post about cyber-physical systems.

Can you imagine a futuristic port where cranes, vessels, AGVs, trucks, and containers communicate, and a central intelligence orchestrates all logistics to optimize operation and to reduce waiting time for cargo and trucks for loading and unloading containers?

Or even a hospital that consists of many different elements communicating with each other to achieve the best global result?

These are just a few examples of what cyber-physical systems (CPS) could look like in the future of industry 4.0. But the uses and roles CPS can play are almost endless.

In fact, cyber-physical systems is one of the five key technology trends that Ericsson CTO Erik Ekudden envisions delivering truly intuitive interaction between humans and machines.

The importance of cyber-physical systems

CPSs are integrations of computation, networking, and physical processes: the combination of several systems of different nature whose main purpose is to control a physical process and, through feedback, adapt itself to new conditions, in real time.

CPSs are transforming the way humans interact with engineered systems, just as the internet has transformed the way people interact with information. Humans will remain crucial in this scenario. As the most flexible and intelligent “entity” in the CPS, humans assume the role of a sort of “highest-level controlling instance”, supervising the operation of the mostly automated and self-organizing processes.

A CPS, being composed of many heterogeneous elements, requires complex models to define each sub-system and its behavior. Dynamic interactions among sub-systems are then orchestrated by an overarching model: a control entity which ensures a deterministic behavior of each sub-system. Current design tools need to be upgraded to consider the interactions between the various sub-systems, their interfaces, and abstractions.

Communication performance in terms of latency, bandwidth and reliability largely impact the dynamic interactions between sub-systems. For a wireless network, factors like device location, propagation conditions, and traffic load change over time. This means that the communication network also needs to be integrated as one of the models in the overall CPS “model of models”.

The time it takes to perform a control task may be critical to enable a correct functioning system. Physical processes are compositions of many things occurring in parallel. A model of time, that is consistent with the realities of time measurement and time synchronization, need to be standardized across all models.

In the future, CPSs will be present in all industry sectors and, within the Industry 4.0 paradigm. CPSs will open new production methodologies becoming the standard of tomorrow for industry. Production environments will be self-configuring, self-adjusting, and self-optimizing, leading to greater agility, flexibility, and cost effectiveness. As illustrated below, every functional aspect of a production chain will be affected, from design, to manufacturing, through supply chains, and extending to customer service and support.

Manufacturing plant with CPS

In the end, the future factory will be a CPS, or a set of interacting CPSs, where highly skilled workers will have insights of the operations directly from coordinated intelligent machines and from a central control entity. This factory will be hyper-connected and data intensive, based on a 100 percent secure industry-grade 5G network.

The pace and approach at which companies adopt digital transformation will vary. However, the step-by-step deployment path towards the full Industry 4.0 paradigm, is unique and well defined. It consists of six stages, each building on the previous one, as illustrated below.

The Industry 4.0 maturity index

On this path, the first two steps are computerization and connectivity. A complete integration of computation, networking, and physical processes has not yet been achieved at this stage. In a third step, with the use of sensors and digital models, the visibility of technology records and processes is achieved in real time. Digital twin models are part of the game here.

With digital models, industries not only see what happens, but also detect and understand, reaching the fourth step of transparency. Here it’s necessary to distil and interpret data by comprehensive data analyses, at all factory levels.

The forecasting ability and predictability is the fifth step. It is about creating different scenarios, estimating the probability of occurrence, and being ready for the possible consequences. The digital models are then dynamically updated.

The Industry 4.0, however, will be fully implemented with the sixth and last step: adaptability. It is about automatically taking adjustment measures without delay and, when needed, without human intervention.

To address all the above, research and development in areas like radio networks, cloud and machine intelligence will be fundamental to realizing the full potential of CPSs.

Since 2016, Ericsson has partnered with Comau, a world leader in industrial automation. Together we have experimented with the steps required to build the factory of the future in a real industrial context. More specifically, we have explored low-latency network and edge cloud for manufacturing plants. Details about these experiments, including the results, were published in Ericsson Technology Review earlier this year called: “Industrial automation enabled by robotics, machine intelligence and 5G”.

The Ericsson-Comau proof-of-concept in Torino, Italy.

The joint research with Comau addresses the evolution towards Industry 4.0. The critical requirements of wireless networking in a plant have been the main challenge, as industrial protocols demand tight radio performances with high reliability, availability, and security. Wireless connected sensors (pressure, temperature, vibrations, cameras) feed expert systems to determine what is happening, to predict future issues, and to prevent failures. Adaptability is addressed by enabling and implementing the centralization and virtualization of controllers, at all levels: plant, production lines, working robotic cells, single robots, actuators. Having a central, global control entity facilitates a fast reaction and an autonomous response to any event and, when needed, to invoke the action and decision of the human supervisor.

Planned next steps are i) implementing the remote control of AGVs and, in perspective, swarms of coordinated AGV concurring to accomplish the same task, ii) testing connected exoskeletons to offload workers while doing their job, iii) zero-touch set-up and operation of production island (if a new robotized island is included in the factory, it configures itself in order to be recognized and supervised by the central intelligence).

This closes the circle with CPS: computation, networking, and physical processes, under a single orchestration to bring in future factories a granted autonomy, supervised by humans, and to respond quickly and effectively to changing customer and market demands.

Find out more about what Ericsson CTO Erik Ekudden has to say about the highly adaptable cyber-physical systems in his Ericsson Technology Review article: Five technology trends augmenting the connected society.