Industry 4.0: Waiting for Godot
Waiting for Godot? No! Just another debate on Industry 4.0
Bluevine Consulting invited me to speak at one of their Future 2-to-5 conferences on the Industry 4.0, on June 5th. My participation consisted of a keynote and as panelist in a debate. During the panel discussion, moderated by Artur Olejniczak, I exchanged views with Ralph Talmont and Chris Wróbel, on few hot topics on technology, the short-term future and – of course – what our actions should be, to get ready for it. Based on this particular panel, few considerations – either discussed during the session or derivative reflections from the debate itself – occurred to me and in no particular order, here they come:
- Waiting for Godot: I have a hunch that there is a generalized feeling – among company executives – that they can wait to embrace Industry 4.0, until some sort of standardization process occurs. In other words, many are waiting for a dominant design to emerge, like the one they expect to see rising anytime soon in consumer IOT. And while I agree that in the consumer context, that is a wise approach, I am surprised that in the industrial landscape, where today there is no common industrial / automation standard within a sector, one should expect a dominant design to appear during the fourth wave of industrialization. In particular one of the key drivers of this transformation is the emergence of Open APIs, machine-to-machine communication protocols, which make the traditional notion of dominant design obsolete. Why my ERP should look like the one of my suppliers, if we have data sharing capabilities and translation protocols? What do we need a dominant design for? And in a pure Beckett style, there is a subtle comic dimension to this main tragic aspect: some of these algorithms, notions of ambient intelligence, concepts and protocols, have been existing for a good decade, but it’s now we have the IT infrastructure and the computing power to make them happen: still, are we Waiting for Godot?!?
- We are so obsessed with how the future will look like, that we tend to forget that it is already at the door knocking. A good friend told me that his daughter graduated two years ago from a school in London, and nearly half of her class ended up working in FinTech. An industry, which did not exist when they started their study program. Predicting the future, it’s hard when the patterns of the past are repeating quite predictably in the future. It’s nearly impossible to predict the Future when there is no relation between past patters and future occurrences. In addition to that, we all should be used by now to foresight, where the focus is almost entirely on the journey not the destination. When we look at the future, we look at multiple scenarios – by focusing on the most important vectors. Never should the foresight exercise lead to one possible future. Yet we are obsessed with this notion of predicting the future: while the future is here already. Telematics is here, now; platooning is a reality. Autonomous driving is a reality. Customers and Consumers are – often – not ready to cross the chasm, and government and regulators are discussing while pretending that there future has not happened already. Technology, though, is here already.
- If we want to prepare for the future, let’s revisit our approach to education. Ralph Talmont made an engaging argument about the challenge of an educational system built on old patters, preparing people for functions that are ceasing to exist. My takeaway is simple: being not ready for cyber-physical system in industry, can destroy more jobs than cyber-physical systems taking over traditional manual labor. The lack of readiness of our educational system is posing a higher unemployment risk than technological evolution itself.
- Big data sounds heavy, so we won’t really need to keep all that data. Like in medical research, AI algorithms sorting through big data can identify relevant patterns and correlations, which are humanly impossible to distinguish; the link between big data and AI in logistics and industrial context can identify operational opportunities, which we do not know they exist. And as we do not know they exist, we assume they don’t, and thus we do not expect large data sets to prove any useful. How do we help companies solving this conundrum?
‘Til next time, while Waiting for Godot!