Duration: May 2019 – May 2022; Coordinator: John Reiner (Infineon Technologies AG, Germany); ARCES Principal Investigator: Prof. Luca Benini.


The mission of the AI4DI project is to make Europe the leader in Silicon-born-Artificial Intelligence for accelerated edge processing. This means to bring the Artificial Intelligence from the cloud to the edge while making it resilient, safe, and secure for the manufacturing and process technologies of the future.

AI4DI Objectives:

  • Accelerate the Artificial Intelligence (AI) adaptation to serve European priorities in the digitizing of the industry

In particular, this means fostering the application of Artificial Intelligence in all European industries in order to gain profit from the efficiency increases, extended knowledge and new forms of control, organization, and business models. Europe urgently needs to take the chances AI brings to synergistically use it with intrinsic strength of being able to create, develop, and manufacture highly complex products.

  • Maximize the benefits of Moore’s Law and More Moore, and revive Moore’s Law beyond the current technology

AI gives the chance to be able to create processes and products, which follow the Maxime of Moore’s Law, and furthermore AI can even drive the semiconductor industry to a level of More Moore. In a time where the Si technologies progress very fast towards its physical boards AI offers, the chance to create new products with vastly increased functionalities, especially by the transition of AI towards the edge devices. Dedicated sensors with hardware integrated AI functions as well as dedicated central processing units with hardware integrated AI functions can transform the way people as well as the industry perceives computational power since these devices offer the chance to integrate intelligence for enhanced sensing, improved collaboration and complex self-organizing control systems which can react more flexible than today’s deterministic systems. In addition, significant performance increases are realizable since these devices do not rely on conventional analytical models, which can become very time consuming, especially when processing multi-parameter-sets.

  • Build a European AI community with European values

AI can be perceived as a tool for solving certain tasks. Thereby many people believe it is necessary to use AI like every other mathematical tool they have adapted to over years. However, the usage and implementation strategies are certainly much more different than people are used to. AI methods are much more flexible and faster to build than most people believe. In particular the flexibility AI methods offer is new to most users. Once a powerful neural network is configured, the only difference between being able to identify/cluster everyday objects and integrated circuits is the training data which has been utilized. Knowing that difference to classical model-based approaches makes it possible to exchange powerful methods/tools in between different industries. Thus, AI4DI wants to foster the collaboration among the industries and academia in order to utilize new findings faster between industries in order to be.

ARCES participates to the project as third part of IU.NET consortium.