Sergio Dulio – Companies have become familiar with the concept of Industry 4.0, the fourth stage in the evolutionary path that, starting at the dawn of industrialisation, transformed manufacturing into the digital process that it is today. It has been a path with important milestones spread over more than two centuries, with huge transformations made possible by specific technologies and new sources of energy.

In 1784 the invention of the mechanical loom fostered the mechanisation of manufacturing. In this first revolution, technology manufacturing no longer needed to rely on physical force whether human or animal, but on new energy sources in the form of water and steam.

Ninety years later, the second industrial revolution opened the door to mass production. Technology multiplied the scale of operations, electricity expanded market dimensions and made available a form of energy that could be easily transported and delivered. The appearance in 1870 of the first assembly lines, which later inspired Henry Ford in the realisation of his highly efficient automobile factories, allowed for new forms of work organisation suited to sustaining large scale production. Fast forward to the 20th century and the manufacturing paradigm changes for the third time.

1969 is the year of the conquest of the moon together with the advent of electronics, computers and automation. With this third revolution, technology multiplied data processing speed, allowing the first substitution of human workers with ‘automata’, machines capable of executing their tasks automatically and unattended.

We are now in the 21st century with Industry 4.0, cyber physical systems, Internet of things and smart technologies. With this fourth revolution, technology multiplied the amount of data that machines and systems can generate and the extent of their use. There is as yet no official acknowledgement of when the term Industry 4.0 was coined but it is widely accepted that the first mentions were made in Germany in 2011. So, it looks as if time was accelerating with less than 50 years passing between the third and the fourth of these crucial milestones.

Now, after ten years, a new evolution is approaching, a new phase that someone started calling Industry 5.0. Are we perhaps approaching what the american inventor, futurist and author Ray Kurzweil calls a ‘technological singularity’, an era where the technical progress advances at an unprecedented pace that is beyond our imagination?  No answer yet exists, but things are certainly changing fast and often unpredictably.

Like the other steps in this journey, Industry 5.0 is enabled by specific technologies. If on the one hand the new paradigm is built on many of the elements we find in the portfolio of 4.0, such as robotics, cloud computing, additive manufacturing and augmented reality, it is as if three in particular are giving a new boost to all of them: collaborative robotics, fifth generation (5G) communication networks and Artificial Intelligence (AI). These are concepts that are becoming ever more common in our regular conversations. It is nevertheless worth examining them in some detail to understand how powerful their effect could potentially be on the next transformation of how we think and produce all our manufactures. So, let’s learn more about them and see how their combination with the newest generations of collaborative robots will again transform matters.

A new kind of collaborative robots

Advanced robotics is one of the underlying technologies in Industry 4.0 and a main element of the advanced manufacturing solutions that are intrinsic in the 4.0 paradigm. These are based on the widespread use of advanced, interconnected and easy to program robotic systems. The most important difference with respect to the more conventional concepts of robotics, is the man–machine collaboration that allows human operators and robots to share the same space and to contribute to the main task. In a typical scenario, tasks of high manual and cognitive complexity are typically assigned to the operator, whilst the robot servant would take care of the simpler and more repetitive ones. This is what led to the development of collaborative robots that are taught exactly how to work in these situations.

Cobots, as they are known, have a mechanical structure that is very similar to that of their conventional brothers but may also have an additional rotational axis to achieve a higher level of dexterity. Cobots are lighter and intrinsically safer to allow the close proximity of operators. So far nothing new. This breed of robots appeared almost ten years ago and it became a fundamental element in the transition towards digital manufacturing and Industry 4.0. At the beginning of this path, these machines were very much like the conventional industrial ones. They were senseless and blind, so they had no means of assessing the environment around them and to let their actions be guided by their perception of it.

What we have seen in the last generations of these machines is exactly this. Robots are developing ‘senses’ through which they are aware of what they are doing and which they can use to drive their actions. They are being covered with sensory skins that can tell them even before a contact occurs that they are getting too close to an obstacle or to a human worker. More recently, they are being equipped with ‘eyes’ in the form of digital cameras that allow them to see and react to the external situation. This is the kind of transition that is shifting us from 4.0 to 5.0. If you can feel and see, you need ‘intelligence’ to read these stimuli and decide how to act. Welcome to the world of intelligent robotics.

AI and machine learning

AI refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. The term may also be applied to any machine that exhibits traits associated with a human mind such as learning and problem-solving. The ideal characteristic of AI is its ability to rationalise and take actions that have the best chance of achieving a specific goal. A subset of artificial intelligence is machine learning, which refers to the concept that computer programs can automatically learn from and adapt to new data without being assisted by humans. Deep learning techniques enable this automatic learning through the absorption of huge amounts of unstructured data such as text, images or videos.

Machine learning (ML) is therefore one of the areas in which the broad domain of AI is structured. ML algorithms are designed to analyse large amounts of data and to identify repetitive patterns in these data from which to infer choices or decisions. The software can then ’learn’ from these data without the need to be specifically programmed for each individual task. Think of a robot that, instead of being programmed for a task, can learn spontaneously how to execute it by watching its human ‘trainer’ performing it for the first time or that decides by itself the best route to take to pick an object from point A and deposit it in point B, avoiding obstacles and choosing the fastest possible path. A clear leap forward with respect to the complex and tedious programming that is required with conventional robots.

We must not think that this is too far-fetched or futuristic. AI and ML are already here in the search engines we use in our computers, in the speech recognition systems of our smart phones, in facial identification systems and in self-driving vehicles. We are again on the verge of another major paradigm shift, a singularity in the course of technological evolution. In this respect, Sundar Pichai (CEO of Google) described some time ago the implications of the development of AI and ML as being deeper than the discovery of fire or electricity. Sufficient to say, we are ready for Industry 5.0 where artificial intelligence and robotics marry to give robots the ability to see, understand and decide.

5G telecommunication networks

It is interesting to see how the development of the cellular networks that we all know and use in our everyday lives went through five evolutionary phases just like the five industrial paradigms described earlier. The difference is that for manufacturing the evolution spanned more than two centuries. In the case of mobile telecommunication technologies it unfolded in roughly 40 years. The first-generation 1G mobile networks   appeared in 1981 and they were for voice calls only.

Ten years later (1991) 2G came along and with it the possibility of sending short text messages (SMS). 3G in 1998 allowed the full fruition and sharing of multimedia contents and access to the Internet. In a quest for more speed and more data to transfer, 4G in 2008 allowed for the first time a true possibility of mobile data consumption and now 5G, first introduced in 2019, is the core technology for the Internet of things and robotics. What this story illustrates is a progressive move from voice to data. Today, cellular networks carry more information (images, videos, files) than voice communications. Bandwidth, speed and low latency are the key elements to enable this change.

Speed measures how many data the network can transmit in a unit of time (typically in megabits per second). Bandwidth indicates how many channels the network can manage at the same time. The higher the number, the lower the risk of overcrowding and lack of response from the network. And, finally, latency, the time interval between an input signal and the arrival of the relative output. The lower the latency the more responsive the network. Future 5G networks (higher bandwidth, higher speed and lower latency) will  enable us to use a number of previously unthinkable services:  5G factories of tomorrow will use the mobile network to exchange low-level signals and control data thus allowing  great deployment flexibility.

Transition of 5.0

How are advanced robots, AI and 5G networks acting together to foster the transition of Industry 5.0? Robots play a more and more important role in innovating and optimising processes not only in manufacturing, but also in agriculture, healthcare and many other industries. For example, mobile robots such as quadrupeds, rovers and drones which can autonomously move around, can support or even substitute human tasks and increase operational efficiency and safety. Today their huge potential is limited, in most applications, by the need for local control or at least local supervision by a human operator. Enabling human supervision or control from a remote centre, where a qualified operator can supervise up to one hundred robots, will exploit their full potential and move us into the cloud robotic era.

While cloud robotics platforms are coming from different players, we still lack the kind of connectivity needed for a professional remote control of a robotic fleet. That is where 5G will fill the gap. Its imminent release will implement key features that will provide the connectivity needed for cloud robotics. The latter also needs to be executed from the edge of the network, both to reduce latency to the lowest level and to avoid any possible disconnection or bottlenecks. Furthermore, robots working remotely will send potentially huge amounts of data from the field, videos or thermal cameras, from laser scans and so forth. All these data must be analysed automatically from AI algorithms, sometimes to react in real-time. The more these algorithms are hosted on the edge of the network, the more they will be efficient and effective in promptly solving dangerous situations and will decongest the rest of the network.

One of the most important challenges of the constant monitoring of the health of the endpoints of these networks will be solved by the low latency of 5G networks. Furthermore, their ultra-broadband allows a robot to be controlled remotely, where it is essential to have a clear vision of the surroundings from high definition, possibly 360° cameras. 5G-controlled robots will be increasingly pervasive and the growth of 5G combined with 5G Mobile Edge Computing (MEC) and AI-based, real-time applications, will enable the creation of a true ‘robot ecosystem’, comprising fixed as well as mobile robots. Smart Robot-based businesses will expand 5G-based hyper-connected services and secure leading technology in edge cloud over the next few years. If you think this scenario is not mind boggling enough, just consider that 6G is already in the pipeline. So the future, as usual, will surprise us.

CREDIT: SHUTTERSTOCK / OLIVIER LE MOAL

Share this story: