Mr. Nachiappan Chockalingam

Privacy-Preserving Federated Learning for Industrial IoT: Enabling Secure Predictive Maintenance Without Centralizing Sensitive Data

Abstract:

In today's Industry 4.0 landscape, leveraging distributed sensor data for predictive maintenance is critical for reducing downtime and operational costs. However, traditional centralized machine learning approaches create fundamental conflicts between AI effectiveness and data privacy forcing organizations to choose between competitive intelligence protection and predictive capabilities. This session explores how Federated Learning architectures enable collaborative model training across distributed IIoT devices while ensuring raw operational data never leaves local premises.

Attendees will gain insights into a three-tier hierarchical framework that transforms privacy-preserving AI from theoretical concept to practical deployment. The talk will walk through real-world implementation on turbofan engine failure prediction, demonstrating how local device training, edge-tier aggregation, and cloud-level coordination achieves similar performance without compromising the privacy of the systems.

Key topics include: Privacy-preserving machine learning fundamentals for industrial applications; Hierarchical federated architectures for resource-constrained IoT networks; Adaptive model pruning and heterogeneity-aware training strategies; Addressing class imbalance and non-IID data in federated settings; Regulatory compliance and data sovereignty in AI deployments.

This session is ideal for industrial IoT architects, data scientists, privacy officers, and manufacturing technology leaders seeking to deploy AI-driven predictive maintenance while maintaining complete data confidentiality and regulatory compliance.

Profile:

Nachiappan Chockalingam is a software engineering leader at Meta Inc, where he architects cutting-edge privacy infrastructure solutions for the Monetization organization. He spearheads initiatives that leverage Artificial Intelligence to navigate the complexities of evolving global privacy regulations, ensuring Meta's infrastructure remains adaptive and compliant with current and emerging requirements.

As a technical leader, Nachi directs cross-functional teams in building privacy-aware systems that safeguard user data and uphold consent frameworks, preventing data leaks while enabling business-critical operations. His work sits at the intersection of scalable infrastructure, AI innovation, and regulatory compliance addressing one of the tech industry's most pressing challenges.

Beyond Meta, Nachi serves as a Senior Member of IEEE, actively advancing research in privacy-preserving AI across diverse domains, including predictive maintenance in Industrial IoT and cloud data privacy. He is a recognized voice in the field, contributing research and insights at international conferences focused on AI ethics and data privacy.

A committed mentor, Nachi has guided numerous junior and mid-level engineers and product professionals, helping them navigate their careers and achieve both technical excellence and professional growth.