FedKit: Enabling Cross-Platform Federated Learning for Android and iOS

TL;DR

FedKit tackles the practical barriers to cross-platform federated learning on Android and iOS by introducing a complete pipeline that converts Python models to platform-native formats, provides unified training APIs, and enables cross-platform aggregation. The system is complemented by backend-driven MLOps that decouples model delivery from app updates and supports multiple concurrent FL sessions via Flower. The authors demonstrate FedKit's viability through a real-world FedCampus health-data use case and a model deployment demo on Android/iOS, showing accelerated local training and improved training loss. The work offers a scalable, production-oriented framework for mobile FL that reduces development friction and enables rapid experimentation across platforms.

Abstract

We present FedKit, a federated learning (FL) system tailored for cross-platform FL research on Android and iOS devices. FedKit pipelines cross-platform FL development by enabling model conversion, hardware-accelerated training, and cross-platform model aggregation. Our FL workflow supports flexible machine learning operations (MLOps) in production, facilitating continuous model delivery and training. We have deployed FedKit in a real-world use case for health data analysis on university campuses, demonstrating its effectiveness. FedKit is open-source at https://github.com/FedCampus/FedKit.

Figures (3)

• Figure 1: FedKit Model Pipeline for Cross-Platform FL.
• Figure 2: FedKit FL Workflow.
• Figure 3: FedCampus Experiment Setup on Duke Kunshan University Campus.