The promise of artificial intelligence often centers on code and cloud computing, but in the high-tech industry a gap remains between a smart algorithm and a working machine. Companies struggle to integrate data-driven intelligence into complex mechatronics, often because of data shortages or hardware limitations. The solution does not lie in software alone, but in integrated system design.
Demcon bridges exactly this gap. Demcon is a group of specialized technology companies that develop, manufacture, and deliver high-end systems and innovative products. Its structure of multiple smaller entities enables the organization to leverage the collective knowledge of 1,200 colleagues without losing the agility and speed that often stand in the way of larger enterprises.
Its expertise ranges from high-tech and AI & data to life sciences & health, quantum, photonics, maritime, and energy. Demcon’s strength lies primarily in its multidisciplinary approach, in which disciplines such as mechatronics, mechanics, software, electronics, and physics come together in a single integrated system design. Barend Vermeulen, managing director of Demcon high-tech systems, and David Rijlaarsdam, managing director of Demcon data driven solutions, offer a glimpse into this synergy.
an integral part.
For Demcon, AI is not a standalone technology, but an integral part of the entire development process. Vermeulen emphasizes that the challenge in mechatronics often lies in the need for lightning-fast feedback loops. “The faster you want to work, the more bandwidth and data transfer you need.” It is precisely these growing data streams that are increasingly becoming a limiting factor. AI can help select or reduce data earlier in the process, allowing systems to continue operating faster and more accurately. Depending on the application, an algorithm can analyze this data locally and in real time via edge computing, turning it into directly actionable insights.
Beyond AI, system design is also playing an increasingly prominent role. Modern mechatronic systems are so advanced that the interplay between all interactions and parameters is becoming ever harder to oversee. Vermeulen: “How do you ensure that all those relationships and interactions remain manageable?” The challenge lies primarily in keeping that complexity under control throughout the entire development process and maintaining visibility into the consequences of changes within a system.
data as the starting point.
For many organizations, AI begins with an algorithm, but according to Rijlaarsdam, the biggest challenge often lies earlier in the process: the data. “Before you even start on an algorithm, the key question is: is there enough usable data to develop it?” Demcon therefore supports clients across the complete data cycle: from acquiring, understanding, and enriching data to integrating it into system design, software development, implementation on the machine, and lifecycle management of the system.
Rijlaarsdam notes that the reality of AI is often complex. He draws an essential distinction between general AI and the specialized machine learning that Demcon focuses on. When data is missing, cannot be used due to privacy legislation, or does not provide sufficient coverage for certain use cases, synthetic data offers a solution. “We also generate that data ourselves, which can really accelerate a project.”
from model to machine.
This integrated approach prevents silo formation, where software and hardware fail to align. Vermeulen sees digital models primarily as a way to accelerate development: software and algorithms can already be developed and tested before physical hardware is available. As the system takes shape, these models are supplemented or replaced by real components, allowing different development steps to proceed in parallel.
Moreover, implementing AI is not purely a technical matter: in heavily regulated sectors, market access often depends on compliance. “It’s not just about the technical dimension, but also the regulatory one,” says Rijlaarsdam. Demcon therefore considers regulations and compliance from an early stage, ensuring that technical innovations are actually viable in the intended market. In a geopolitically unstable world, sovereignty is also playing a growing role. Companies want full control over their intellectual property and data. Vermeulen: “We can do all of that on-premise, which means security and sovereignty can be kept very high, depending on the client’s requirements.”
knowledge transfer.
Given the high degree of confidentiality in these projects, knowledge and intellectual property are always important topics in conversations with clients. “We work with a collaborative model,” Vermeulen explains. Such partnerships involve a long-term relationship, in which the developed technology ultimately belongs entirely to the client, while the accumulated experience and expertise remain within Demcon. “Successful collaboration on complex projects like these is impossible without trust.”
