TechEx Europe in Amsterdam brought together big names from a wide range of industries to discuss ongoing digital projects covering business transformation, manufacturing processes and the ongoing adoption of IoT, among others.

Undoubtably the key theme throughout the event was a familiar one for anyone covering the enterprise digital revolution: the quality of data is key.

Whether a generative AI application or predictive maintenance for a factory floor, the information going in is perhaps the most critical element if you want to get something usable out.

However, there were also some tips and insights that are perhaps less obvious and could only be garnered from real world experience. Here, Mobile World Live rounds-up a quartet of top takes from those on the ground.

Digital twins are no silver bullet
Some of the marketing around digital twin adoption has been almost evangelical in nature of late and, although there are certainly benefits of using them in the right setting, Volvo Group IoT expert Julien Bertolini cautioned of the importance of being realistic and targeted in their use.

During a session dedicated to the technology, Bertolini emphasised: “Take your time to have a clear definition at the beginning. Think big start small, you should always start like that”.

“The most important thing to keep in mind is it should be related to a use case. If you have a clear business pain point to solve the digital twin will help you. If you think you can use it to solve all your issues it will be a failure for sure”.

AI isn’t always right
Among the overpowering hype around AI seen across the event, the value of people in processes was demonstrated by Alessandro Malucelli, industrial engineering senior manager for domestic appliance manufacturer Whirlpool.

A man in a suit is speaking into a microphone on a stage labeled

During a session on the planning and building of its smart factory in Argentina, Malucelli (pictured, left) said an AI model “assumed we could cover the top of the roof with solar panels, which could capture solar energy to feed the lighting”, in an attempt to up sustainability of the facility.

However, questioning this, he indicated a better solution was a design incorporating “top to bottom” natural light instead.

This was one example of how human intervention had improved various aspects of the factory.

He noted, however, had the company had “a lot more of money”, it could have ended up with far less people-input, suggesting it may have just used data and robots. This would mean losing insight from those with real-world experience able to change the design and improve the factory floor.

Lined-up data sources are a must
Problems associated with data silos and formatting were brought up numerous times by speakers, with the value of getting the whole picture deemed vital for accurately assessing the performance of facilities and enterprises as a whole.

The point was emphasised by Airbus manufacturing process data acquisition and processing expert Alejandro de la Mata, who said “the benefits are found immediately” when data is standardised across an organisation.

“If you have one single source of truth, you have a clear source of information so you can quickly investigate in your [current] data, and your historical data” of a manufacturing process. “You can anticipate problems, or you can look at what is happening in the event you have issues”, he added.

IoT is not Field of Dreams
As the mis-quote from the film Field of Dreams goes, “if you build it, they will come”, however this mantra does not always make for a success in the technology sector.

The need for a specific use case for innovations was raised by multiple speakers.

A person stands in front of a blue background with

In the realm of IoT, Henkel Adhesive Technologies chief technology lead and head of IoT competence and enablement Philipp Mahr (pictured, left) noted when building an IoT proposition, it is vital to have a potential customer on board from day one.

“Building an IoT business is not easy, there are always challenges. Technology isn’t always the hardest part,” he said. “Don’t start anything without a customer, you will learn fast and will spend less”.

“Try and charge them from day one,” he added, but highlighted this fee could come in the form of investing time in the project rather than hard cash as “customer involvement is actually proof you have some interest.”