Assess learners' ability to design a scalable, privacy-preserving federated learning system for heterogeneous edge devices by specifying architecture, selecting and justifying differential privacy mechanisms (local vs. central DP, noise calibration, privacy accounting), choosing a secure aggregation approach (MPC, HE, or hybrid), proposing straggler mitigation and client-selection strategies (asynchrony, deadlines, partial updates, coded computation), addressing communication/computation compression and personalization for device heterogeneity, and outlining an evaluation plan that measures privacy–utility trade-offs, latency, robustness, and scalability with concrete experiments and metrics.