Closing the Edge AI Gap: From Model Training to Real-World Deployment

Wednesday, June 10, 2026
10:00 a.m. - 11:00 a.m. Central Time

Connect model training, edge deployment and cloud-based lifecycle management for real-time optical inspection

Hosted by:

IMC-Logo-clear.png

Register for this webinar
Webinar

Presenters:

Andreas Burghart
Andreas Burghart
Senior Product Manager
Digi International
Daniel Amor
Daniel Amor
Innovation Manager
RBZ Robot Design

Industrial teams continue to face challenges when moving from AI model development to real-world deployment at the edge. From training and optimizing models to deploying them across different hardware accelerators, the process can slow time to market and increase system complexity. At the same time, traditional inspection systems often rely on expensive, centralized computing that limits scalability and flexibility.

This webinar explores how organizations can move from AI model development to real-world deployment and ongoing optimization at the edge. Using a real-world optical inspection example from food production, the session demonstrates how edge AI and cloud-based services work together to simplify the full lifecycle, from training and deployment to monitoring and continuous improvement.

Presenters will walk through the complete workflow, including how to deploy models on embedded platforms, manage performance across accelerators and maintain systems over time through cloud-based updates and monitoring. The session also highlights how cloud services support continuous improvement through model retraining, over-the-air updates and system visibility.

Attendees will learn how to:

  • Deploy edge AI for real-time optical inspection in industrial environments
  • Manage model training, optimization and deployment across multiple accelerators
  • Use cloud infrastructure to monitor devices and update models over time
  • Reduce system cost and complexity compared to traditional inspection approaches
  • Scale AI applications across use cases while maintaining performance and flexibility