The HiCONNECTS project (Heterogeneous Integration for Connectivity and Sustainability), funded through HORIZON-KDT-JU-2021-1-IA with GA 101097296, is a cutting-edge initiative aimed at revolutionizing the current cloud computing landscape. Its main goal is to develop heterogenous integration (HI) core technology solutions that can address two major societal and industrial challenges: the transmission of Internet of Things (IoT) data over the IT network and the sensing of objects to enable Highly Automated Driving (HAD). TO achieve the objectives, the project brings together a consortium of 64 partners (large industrial players, universities, RTOs, and SMEs) The focus of HiCONNECTS is to develop sustainable and energy-efficient ways of transforming the centralized cloud platform into a decentralized edge cloud computing system. This approach will bring cloud services, including Artificial Intelligence (AI), closer to IoT end-users. The expected results of the project include an increased ability to move mobile clients between different places with minimum cost, short response time, and a stable connection between cloud nodes and mobile devices.
PoliMi, represented by the team of researchers from the Department of Mechanical Engineering, is playing a crucial role in the HiCONNECTS project working in a close collaboration with ST Microelectronics Italy and with the Università di Catania. They are designing an AI-Enhanced Digital Twin architecture for the multi-fidelity prediction of semiconductors facility performance. PoliMi's contribution involves defining the models and methods needed to develop and operate digital twins that represent the facility, as well as creating AI-based multi-fidelity models that use both high-fidelity simulations and real-time data. The objective of the architecture is to improve production performance by increasing production throughput, optimizing machine utilization, and reducing lead times. Indeed, in semiconductor facilities nowadays, the existing solutions for dispatching wafer-lots across equipment are able only to assign priority to lots by simple rule-based heuristics, with the aim of enhancing the performance in terms of Work In Process (WIP) level and throughput related targets. The PoliMi goal is to exploit the novel AI-Enhanced Digital Twin to allocate manufacturing resources to wafer lots and then to dispatch those through the machines, based on data coming for the physical production environment.
In conclusion, HiCONNECTS project, along the researchers from the Department of Mechanical Engineering in Politecnico di Milano led by Professor Andrea Matta, is pushing the boundaries of cloud computing and IoT technology to drive innovation, enhance performance, and bring cutting-edge solutions to major societal and industrial challenges.