Battery Range Prediction using Federated Learning on Edge
summary
Uses a simulator to generate time series of uphill/downhill and auxiliary-load scenarios and trains a PyTorch predictive model locally on constrained ECUs, then shares model updates (not raw data) to a central aggregator using the Flower FL framework over gRPC. Central aggregation runs containerized on-prem or cloud, edge inference and retraining run as a quality-managed component targeting a Texas Instruments platform. Claims practical edge training and deployment workflow from simulator to aggregated model for battery range prediction.
AI-written summary · about
chapters (7)
- 0:00Federated learning at the edge
- 1:28EV range anxiety methods
- 3:53Battery drain simulator data
- 4:56Federated learning reduces latency
- 6:21Flower framework
- 7:46Federated learning architecture
- 10:26Q&A
related talks
48mFederated ML for Edge Applications
13mScalable and Portable Federated Learning Simulation Engine
32mInnovation in the Age of Regulation: Federated Learning with Flower
36mEdgeLake-FL: An Automated Federated Learning Platform for the Edge
28mA Privacy Enhancing Federated Learning Framework for Multi-Party Distributed ..
12mUsing AI-enabled Edge Cloud Technologies to Strengthen Europe’s...
57mKeynote: Federated Learning with Flower: AI's Next Frontier
32mPlant a Touch-Me-Not: Train Models Without Anyone Touching Your Data with Flower