Table 8 Optimal performance of the proposed multi-node edge and cloud computing set-up.
From: Leveraging federated learning and edge computing for pandemic-resilient healthcare
Block | Specific item | Average error at node level | Average error at system level |
|---|---|---|---|
Sensor type | Camera | 4% | 3.80% |
IR thermometer | 5% | 3.50% | |
Proximity sensor | 3% | 2.80% | |
Ultra-sonic | 3% | 2.40% | |
Decision support for facemask | Low (30 person/ hour) | 9.2% | 8.5% |
Medium (50 persons/ hour) | 11% | 8.8% | |
High (70 persons / hour) | 12% | 9.3% | |
Decision support for social distance | Low (30 person/ hour) | 9% | 3.40% |
Medium (50 persons/ hour) | 12% | 4.50% | |
High (70 persons / hour) | 13% | 4.50% | |
Decision support for contact tracing | Low (30 person/ hour) | 8.5% | 3.80% |
Medium (50 persons/ hour) | 10% | 3.40% | |
High (70 persons / hour) | 12% | 4.50% | |
Latency variation (average) | Low (30 person/ hour) | 9% | 7.5% |
Medium (50 persons/ hour) | 14% | 11% | |
High (70 persons / hour) | 15% | 14% |
