Page d'accueil // SnT // News & E... // Next-Generation Industry Automation

Next-Generation Industry Automation

twitter linkedin facebook email this page
Publié le mercredi 24 mai 2023

The conference “Applied Automation in Teaching and Development” (AALE) took place in March 2023 for the first time in Luxembourg, hosted by the Lycée des Arts et Métiers. The event was organised by the BTS Connected Buildings and Cities and the German association VFAALE e.V., which mainly represents universities of applied sciences in Germany. The theme of this 19th  edition of the AALE conference was “Deploying Automation against Climate Change”.

The objective of the conference was to provide a platform for the exchange on automation technologies in light of digital transformation, industry 4.0 and smart buildings. This cross-border event aimed at discussing automation technologies with researchers and industry, but also to introduce high school students to the topic. To this end, the conference offered workshops for high school students to pique their interest in studying engineering sciences.

Sylvain Kubler, research scientist at the University of Luxembourg’s Interdisciplinary Centre for Security, Reliability and Trust (SnT) was a speaker at the conference, where he also held a workshop – with his PhD students  (Dorian Joubaud and Guilain Leduc) – for high school students. We sat down with him to get some insights into his approach to automation technologies and his experience working with high school students. 

Sylvain, you held a session about digital twins at the conference. What are they, and what was your main message? 

A digital twin is a virtual representation of a physical object or process that helps us optimise business performance. Going further, you can simulate future industry processes, foresee what could go wrong, and already improve these processes before implementing them. My main message was that implementing such a digital twin in industry requires time. The company must go step by step, and getting ready usually takes a couple of years.

Can you give us an example?

If you are manufacturing a car, a digital twin would be a sort of avatar of this car, a 3D model of it fed in real time with sensor data. You could use this avatar that is enriched with real-time data to simulate and optimise processes, for example in manufacturing. 

What are the steps that a company must follow to implement a digital twin? 

Companies must first identify use cases for digital twins, then prepare to collect data. However, they cannot just use off-the-shelf machine learning algorithms to analyse this data; they need to refine their strategy to train the algorithms. For this, the next step is to implement a data science lab to have a testbed for the digital twin. In the end, the company can deploy it and monitor its performance over time.

What are your learnings on digital twins from our partnerships with industry?

What we understood is that companies need to set up their own data science teams for the day-to-day work on training the digital twin and collecting data. SnT can help in building these teams through our partnership programme. This is the right foundation to innovate and maintain high levels of innovation and performance in production. And companies need to be aware of the path they need to follow to build a digital twin. It is best to decide case by case if a digital twin is useful. 

What was your workshop for high school students about? 

The workshop aimed at giving the students an idea of Internet of Things (IoT) applications and their development. The students developed an application that provided a service to car drivers to select the best electric vehicle station. It combined, for example, information on weather forecasts, opening hours and the charge capacity of the stations. 


Sylvain Kubler in his workshop for high-school students on IoT applications.


What did the students learn?

Some of the students did not have any previous programming experience, so programming an application was something completely new to them. They learnt that HTTP, which is a protocol with a request-response mechanism used for the world wide web, can also be used to create IoT applications. Indeed, HTTP can be used not only for human-to-machine interactions (what we do on a daily basis using a Web browser), but also for machine-to-machine interactions. The possibilities offered by this approach are open to imagination: we developed a service where we combined information from different platforms in machine-readable format. All students performed well in the exercise, and developed their own application. For some of them, this was the beginning of their interest in programming: they wanted to further develop their service afterwards.