Hardware and Software solutions for edge-cloud AI implementation in High Speed Packaging Machines

Register for the event


In this seminar, we explore a set of hardware and software solutions developed to implement and run efficiently AI algorithms both on-edge and on fog servers.

The target scenario is the one of high speed packaging machines, where the automated machines must transform input materials into desired packages while reducing human workload.

The seminar is thoughtfully structured around four fundamental themes:

  1. Introducing a Novel Hardware Architecture for On-Edge AI Algorithms, that introduces an innovative hardware architecture designed to support on-edge AI algorithms.
  2. Establishing a Robust Data Transmission System in the Fog Environment and Facilitating Concurrent Data Access, that illustrates how to create a dependable system for data transmission within fog environments and enable concurrent data access for multiple processes.
  3. Enhancing Machine Health Status Monitoring, that discusses AI algorithms developed to automatically monitor the health status of the packaging machine.
  4. Enhancing Product Quality Control Processes, that discusses the AI algorithms developed to improve the current product quality control procedures.

The development of the HW/SW solutions that fall in the above themes has been carried out within the framework of Pilot 3 of the European project IMOCO4.E.



10:30 Start of the Seminar

  • Introduction to the IMOCO4.E project and the Pilot 3 – Riccardo Masiero, CRIT
  • The real time smart control platform – Fabien Bouquillon, UNIMORE
  • The TSN platform based on FPGA and the AI operations speeding up module – Daniel Uribe, SoC-E
  • The Huawei’s Atlas 500 Pro – Lorenzo Diana, Huawei
  • The Distributed Messaging System and the Big Data repository – Tassos Kanelos, GNT
  • The Long Short Term Memory AutoEcoder Anomaly Detection algorithm for predictive maintenance – Tassos Kanelos, GNT
  • The AI algorithm for Sensor Anomaly Detection – Dario Guidotti, UNISS
  • The AI algorithm for Image Anomaly Detection  – Abm Tariqul Islam, Digital Twin Technology
  • Q&A

12:30 End of work

Luogo Evento
Ms Teams (Online Event) su Google Maps