This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy

For critical processes, AI is also a danger and should be used very carefully.
04

Why AI's benefits may be over-estimated

A Q&A session with Ulrich Sendler, Author and technology analyst

Q: It has been nine years since your book “Industrie 4.0”, was published. What has and hasn’t changed since that initial vision and what does Industrial Internet of Things mean to you today?

A: Industrie 4.0 was an initiative of the German industry and research under the guidance of the German Academy of Technical Science and Engineering (Acatech), and soon also under the guidance of the government. As I put it in my book, it was about the total digitalization of all processes and methods in the industry, from the idea of a product through engineering, production planning and production to sales and the product use at the customer’s side. One very important aspect was and is the possibility of new services that are possible because of the availability of the Cloud and the chance to connect products as well as users with each other and/or with anything. IPv6, the internet protocol version 6, makes this possible through an almost endless number of Internet connections wi­th a potential IP address for every product, not only for every human being.

We are far away from achieving this. It is still almost only the large companies with thousands of employees that are trying to implement Industry 4.0 standards. Meanwhile, we are talking about the digital twins of products and processes being developed which help to simulate the behavior of products as well as in real life. 

But the vast majority of the industry is still working with traditional methods and processes, using software and internet only in enhancing productivity and velocity. Very few new services have been invented and installed, like predictive maintenance and Product as a Service.

What has changed? The initiative has defined proper standards which are widely used in some industries, mainly in the bigger companies. As I mentioned as my fear in the book, the industry was focusing on the optimization of production efficiency instead of completely changing the processes and finding new business models especially in the field of service.

For the majority of small and medium businesses, it is not the question of having the right software or Cloud provider. It is identifying the new possibilities and business models, based on the very good products that they have and are successful with. The initiative I 4.0 was too much focused on the large industry groups. The SME’s feel left alone.

Q: Industry is not totally digitalized, and you observe that it’s mostly big companies that are benefiting from it. So what can be done to pull – or even push - SMEs into digitalization so that they benefit alongside their larger counterparts?

A: The only way that I see is through networking and sharing their experiences with experts. That’s what I am trying to offer. But the result up to now is very poor.

Q: What has been the impact of the apparent international retreat into silos on the development of Industrial IoT?

A: I think that has not made a big difference to IoT and Industry 4.0. I’m not talking here about the difficulties in accessing the right, well-qualified employees and partners in the field of software development. There were many thousands of people doing this job in Russia and especially in Ukraine, that are not available any more for maybe many years. Even if the political and economic relationships change dramatically, it is still - and even more than ever - the power of the digital, virtual world and the possible connection of every company and every product and service via the Cloud that can enable collaboration across the borders. Although the partners working together are not the same as before.

Q: In your latest book, you discuss the application of AI in industry. What is the potential benefit of AI in the traditional energy industry, for example mining?

A: One important point that I make in this book is that the potentials of AI for the industry are very much over-estimated. It is again more about the optimization of production efficiency than real innovation with new business models and new services. For critical processes, AI is also a danger and should be used very carefully. There is no way to take the responsibility for critical decisions in production away from human beings to AI. AI is principally – and cannot be anything else – statistic average. What you need for critical decisions is exact data and “yes” or “no”, instead of, say, “98.1% yes”. AI cannot deliver that. It is the providers of Cloud platforms and AI microservices that sometimes make it sound like it will be different. And sometimes decision-makers in the industry would like to believe it. That is the over-estimation.

Q: Even if currently over-estimated, what is the potential benefit to industries such as mining of AI and Industry 4.0? Can it help the wider process – not just extraction but the sustainability/life cycle of the fossil fuels themselves?

A: The benefit of using AI lies in the replacing of humans in calculating, remembering and identifying masses of data. That is something the computer is better at, like image recognition.


Contact us! transform@huawei.com