Flow-FL: Data-Driven Federated Learning for Spatio-Temporal Predictions in Multi-Robot Systems

TL;DR

This paper introduces Flow-FL, a novel serverless federated learning approach using gossip-based data sharing for multi-robot spatio-temporal prediction, demonstrating its effectiveness through multi-agent trajectory forecasting experiments.

Contribution

The paper presents Flow-FL, a serverless federated learning variant with a gossip-based aggregation, enabling decentralized learning in multi-robot systems.

Findings

Flow-FL performs comparably to centralized models in trajectory forecasting.

Decentralized data flow impacts learning efficiency and accuracy.

The approach is robust to communication delays and varying data contributions.

Abstract

In this paper, we show how the Federated Learning (FL) framework enables learning collectively from distributed data in connected robot teams. This framework typically works with clients collecting data locally, updating neural network weights of their model, and sending updates to a server for aggregation into a global model. We explore the design space of FL by comparing two variants of this concept. The first variant follows the traditional FL approach in which a server aggregates the local models. In the second variant, that we call Flow-FL, the aggregation process is serverless thanks to the use of a gossip-based shared data structure. In both variants, we use a data-driven mechanism to synchronize the learning process in which robots contribute model updates when they collect sufficient data. We validate our approach with an agent trajectory forecasting problem in a multi-agent setting. Using a centralized implementation as a baseline, we study the effects of staggered online data collection, and variations in data flow, number of participating robots, and time delays introduced by the decentralization of the framework in a multi-robot setting.