Skip to main content
  • Log in
  • Manage Cookies
EclipseCon - Eclipse Foundation
  • Conference
    • Program Schedule
    • Program List
    • Registration
  • Features
    • Community Day
    • Hacker Day
    • Dinner Meetups
    • OSGi Summit
    • Keynotes
  • Community
    • Code of Conduct
    • Health & Safety at EclipseCon
    • Community Resources
    • 2022 Speakers
    • Information for Speakers
  • Sponsors
    • Be a Sponsor
    • Information for Exhibitors
    • Our Sponsors
    • Sponsor Testimonials
  • Venue
    • Conference Venue
    • Hotels
    • Ludwigsburg
  • About Us
    • EclipseCon 2022
    • Program Committee
    • The Eclipse Foundation
    • Past Conferences
    • Other Events
  1. Home
  2. EclipseCon
  3. EclipseCon 2022
  4. Sessions
  5. How to Train Your Dragon and Its Friends: AI on the Edge with Eclipse Kura

How to Train Your Dragon and Its Friends: AI on the Edge with Eclipse Kura

Session details
Status: 
Accepted
Speaker(s): 
Pierantonio Merlino (Eurotech)
Mattia Dal Ben (Eurotech)
Experience level: 
Intermediate
Tags: 
deep learning
Eclipse Kura
EdgeAI
Kapua
AI/ML
Session Track: 
IoT & Edge
Session Type: 
Standard

A case-study of the lifecycle of a machine learning-based application within the Eclipse Kura ecosystem: from data collection to training, deployment and inference on the edge.

We'll focus on the process of creating a deep learning anomaly detector from scratch, leveraging the entire Eclipse Kura ecosystem:

  • Data collection: Kura Wires for collecting diagnostic data from an appliance and upload on the Eclipse Kapua cloud

  • Training: How can we download the collected data from Eclipse Kapua and develop a model within our preferred ML library (Tensorflow/Keras).

  • Inference: How can we export the trained model for running on an Inference Server (Nvidia Triton™ Server) which will be our target runtime.

  • Deployment: How can we deploy this trained model on the edge device leveraging the intuitive Eclipse Kura Wires interface and running the models with Eclipse Kura's inference engine service.

Exciting new and upcoming Eclipse Kura features for Edge AI (AI Model Encryption, New Inference Servers support and more).

Slides: 
PDF icon KuraAI.pdf
Slides (External URL): 
https://github.com/mattdibi/eclipsecon-edgeAI-talk
https://colab.research.google.com/github/mattdibi/eclipsecon-edgeAI-talk/blob/ma…
Objective of the presentation: 
In this talk we will give an introduction of the newly added AI features of Eclipse Kura by showcasing a Deep Learning-based anomaly detection application running on an edge device. This end-to-end example will touch all the main topics of a machine learning deployment: from the data collection to the model deployment to a real device.
Attendee pre-requisites - If none, enter "N/A": 
Basic knowledge of machine learning concepts; some familiarity with Python (not strictly required)
Schedule info
Time: 
25 Oct 2022 - 12:30 to 25 Oct 2022 - 13:05
Room: 
Wilhelm-Krämer-Zimmer
  • Sign in to post comments.

Elite

  • IBM
  • Huawei

Premium

  • Yatta Solutions GmbH
  • Eurotech
  • Red Hat

Basic

  • EclipseSource
  • Obeo
  • Bosch IO GmbH
  • SCANOSS
  • SSI Schaefer IT Solutions GmbH
  • Mercedes-Benz Tech Innovation
  • TypeFox GmbH
  • ARCAD Software

Media

  • GermanTechJobs

Become a Sponsor

Eclipse Foundation

  • About Us
  • Contact Us
  • Sponsor
  • Members
  • Governance
  • Code of Conduct
  • Logo and Artwork
  • Board of Directors
  • Careers

Legal

  • Privacy Policy
  • Terms of Use
  • Copyright Agent
  • Eclipse Public License
  • Legal Resources

Useful Links

  • Report a Bug
  • Documentation
  • How to Contribute
  • Mailing Lists
  • Forums
  • Marketplace
EclipseCon is brought to you by The Eclipse Foundation with the support of our sponsors.
Powered by Drupal and built on COD.

Copyright © Eclipse Foundation. All Rights Reserved.

Back to the top