IEEE Future Directions
Related Articles
A much smoother transition to an Industry 4.0 can be pursued by leveraging the clouds that several industrial players are already using. There are clouds that have been and are used as a way to obtain flexible, on demand processing and storage capabilities (basically lowering the total ownership cost) and there are smaller –private- clouds that companies are using to share data along the supply, distribution chain and that are now progressively used to monitor the usage of their products.
By integrating the various existing clouds, already embedded in the manufacturing processes (extending northbound to the supply chains and southbound to the distribution chains and beyond) a major company may take the lead steering the Digital Transformation from the hub (the company) out (involving its suppliers and distributors). This is what has been happening in the aeronautic manufacturing where Boeing and Airbus are very strong players that can pull their supply and distribution chain in their desired direction.
Other players, like General Electric, delivering high value products (like turbines in the case of General Electric. Notice that the cost of the engines for a 747 or A380 is about one fifth of the cost of the whole plane…) can also have the power to steer the supply and delivery chain in their desired direction and indeed General Electric has its own (private) cloud to keep track of its turbines (more on this later).
Of course, some industries are already operating in the cyberspace (they are dealing with services that are mostly bit based) and companies operating in this space are already using the cloud as their“manufacturing”infrastructure.
Amazon Web Services is now being offered a digital platform and there are specific guidelines in place to enable the Digital Transformation to these Services.
Azure is also offering specific services to manufacturing and support to the Digital Transformation (as well as, obviously, support to all industries already operating part of their business in the cyberspace).
System integrators make use of these large clouds and support companies in the Digital Transformation by integrating local clouds and streamlining all processes to operate in the new environment. The Digital Transformation will make significant use of these Cloud at the edges and connection providers, like Ericsson, are scrambling to integrate distributed clouds (or fog or edge computing) into their connectivity offer.
Indeed, this view of cloud at the edges represent the transition point from the Cloud approach to the Connectivity approach. Companies like Cisco are pushing for this approach.
As a matter of fact, the Digital Transformation is flattening hierarchies: all components in a value chain are potentially interconnected with all the others (once you are moving to bits, accessing them in parallel and from any point of the globe is no longer an issue). Interconnection takes place among organisations, enterprises, suppliers, manufacturers, providers and users, it takes place among products and products components (IoT), it takes place in form of features interoperability aong services … Interconnectivity is the name of the game in Industry 4.0, and it is made possible, conceptually by the shift from atoms to bits and practically by the availability of pervasive connectivity infrastructures.
We have kept saying that communications infrastructures are just killing distance, everything is here and now, but that is true only at macro level. At the level of robots there is still the perception of difference between local parties and distant ones: this is called latency. Some of this latency is connected to the speed of light (actually the propagation of the electromagnetic field…) and this is something that cannot be overcome (although by moving manufacturing component in a cloud one could physically co-locate some activities): controlling a robot in London from Turin implies a delay of 2.7ms per direction so some 5ms if you need to take decisions based on response. The only way around this would be to virtualise the robot and its controller and operate with one another in close proximity in a cloud, but this can be made only if you can virtualise both, and that is seldom the case).
In practice, however, controlling a robot in London from Turin using a plain vanilla telecommunication network is likely to involve a delay (latency) of some hundreds of ms. A radio system like LTE (4G) may on average add a latency between 50 to 100 ms. A properly designed 5G network may significantly cut this latency below 50ms (yes, I know, if you sit on the 5G antenna and no-one else is around to bother you can go down to a few ms, but let’s be realistic…).
Because of this, and the expected penetration of 5G in the long run many consider it as the idea communication infrastructure for Industry 4.0.
In my opinion (shared by many CEO I had the opportunity to talk to in the last year), 5G will be a welcome bonus to manufacturing, once it will be deployed. In the meantime, industry is making the most out of 4G and even 3G.
Nevertheless, it makes sense for Telecom Operators to ride the (future) wave of 5G boasting on its low latency that indeed in some niches could make a difference.
5G is coming with a different network architecture (if it were not the case you should be contented with the present latency…) and in this network architecture you see clouds at the edges becoming an integral component. In this sense a connectivity infrastructure based on an integration of ege computing (cloud at the edges), multi access capability and micro cells delivering high throughput (5G) will be a most interesting platform for Industry 4.0.
However, a pure communications infrastructure, although highly performant and pervasive is not sufficient to be a digital platform enabling Industry 4.0. As I mentioned it has to include the cloud part, which in turns makes sense only if you are including data.
