Enhanced Collaborative Edge Intelligence for Explainable and Transferable Image Recognition in 6G-Aided IIoT | JoVE Visualize

Area of Science:

Artificial Intelligence
Computer Engineering
Telecommunications

Background:

The Industrial Internet of Things (IIoT) integrates interconnected devices and intelligent applications, with sixth-generation cellular networks (6G) enhancing performance for low-latency, high-reliability tasks like image recognition.
Current IIoT image recognition algorithms often lack explainability and fail to exploit the collaborative potential of edge computing.
Addressing these limitations is crucial for advancing AI in industrial settings.

Purpose of the Study:

To propose a novel method, Intelligent Recognition with Collaborative Edges (IRCE), to improve explainability and transferability in 6G-aided IIoT image recognition.
To enhance transparency and trust in IIoT systems through visual explanations of AI decision-making.
To enable effective domain adaptation and collaborative intelligence across distributed edge servers.

Main Methods:

IRCE incorporates an explainable layer into the feature extraction network to generate visual prototypes for decision-making transparency.
Local Maximum Mean Discrepancy (LMMD) loss is integrated to facilitate seamless transfer learning across geographically distributed edge servers.
The method leverages edge intelligence for real-time performance optimization, reduced computational costs, and enhanced scalability.

Main Results:

Simulations show IRCE achieves superior accuracy, explainability, and adaptability compared to traditional methods.
The approach demonstrates efficient operation in diverse environments, highlighting its practical applicability.
IRCE effectively optimizes real-time performance while reducing computational load and improving scalability.

Conclusions:

IRCE represents a significant advancement in scalable, explainable, and transferable AI for IIoT ecosystems.
The method enhances trust and transparency in AI-driven industrial applications.
IRCE's capabilities are well-suited for critical industrial applications like smart manufacturing, remote diagnostics, and intelligent transportation systems.