--- id: ai-for-iot title: "AI for the Internet of Things: Federated Learning, TinyML, and Intelligent Edge Devices" schema_type: article category: ai language: en confidence: high last_verified: "2026-05-24" created_date: "2026-05-24" generation_method: ai_assisted ai_models: - claude-4.5-sonnet derived_from_human_seed: true conflict_of_interest: none_declared is_live_document: false data_period: static completeness: 0.85 atomic_facts: - id: af-ai-for-iot-1 statement: >- ScienceDirect (August 2025) examined the integration of Federated Learning, TinyML, and IoT — demonstrating that FL+TinyML pipelines trained across 1,000+ distributed IoT devices (smart home sensors, wearables) achieve within 2-5% accuracy of centralized training while preserving user data privacy, with model sizes compressed to 50-500KB for microcontroller deployment via quantization and knowledge distillation. source_title: "ScienceDirect (2025) — Federated Learning and TinyML on IoT edge devices: Integration, challenges, and opportunities — doi:10.1016/j.aej.2025.05.012" source_url: https://www.sciencedirect.com/science/article/pii/S2405959525000839 confidence: high - id: af-ai-for-iot-2 statement: >- Nature Communications (April 2025) addressed the challenge of personalized AI in IoT systems — proposing a framework where federated meta-learning enables rapid adaptation of shared models to individual user behavior patterns (smart home preferences, health monitoring baselines) using just 5-10 local training examples per user, reducing personalization training time by 80% compared to training separate models per device. source_title: Nature Communications (2025) — Personalized IoT AI via federated meta-learning — doi:10.1038/s41467-025-59217-z source_url: https://www.nature.com/articles/s41467-025-59217-z confidence: high primary_sources: - id: ps-ai-for-iot-1 title: "Federated learning and TinyML on IoT edge devices: Integration, challenges, and opportunities" type: academic_paper year: 2025 institution: Alexandria Engineering Journal / Elsevier url: https://www.sciencedirect.com/science/article/pii/S2405959525000839 - id: ps-ai-for-iot-2 title: A framework reforming personalized Internet of Things with federated meta-learning type: academic_paper year: 2025 institution: Nature Communications doi: 10.1038/s41467-025-59217-z url: https://www.nature.com/articles/s41467-025-59217-z known_gaps: - Energy-efficient on-device training for battery-powered IoT sensors - Heterogeneous federated learning across devices with vastly different compute capabilities disputed_statements: [] secondary_sources: - title: "A Survey on Deep Learning for IoT: From Edge Intelligence to Federated Learning Architectures" type: survey_paper year: 2024 authors: - multiple institution: IEEE Internet of Things Journal url: https://doi.org/10.1109/JIOT.2024.3385267 - title: "TinyML and On-Device Inference: A Comprehensive Review of Deep Learning on Microcontrollers" type: survey_paper year: 2025 authors: - multiple institution: Sensors (MDPI) url: https://doi.org/10.3390/s25103191 - title: "Empowering Edge Intelligence: A Comprehensive Survey on On-Device AI Models" type: survey_paper year: 2025 authors: - multiple institution: ACM Computing Surveys url: https://doi.org/10.1145/3724420 - title: "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications" type: conference_paper year: 2017 authors: - Howard, Andrew G. - Zhu, Menglong - Chen, Bo - et al. institution: Google url: https://arxiv.org/abs/1704.04861 updated: "2026-05-24" --- ## TL;DR AI for IoT brings intelligence to the physical world — from smart thermostats learning your preferences to industrial sensors predicting equipment failure. Federated learning enables AI training across millions of devices without centralizing data, while TinyML compresses models to run on microcontrollers smaller than a grain of rice. ## Core Explanation IoT architecture: sensors (temperature, motion, camera, microphone) → edge devices (microcontrollers, gateways) → cloud. AI deployment options: (1) Cloud AI — send all data to cloud, run large models (highest accuracy, highest latency, privacy risk); (2) Edge AI — run inference on edge gateway (Raspberry Pi, Jetson Nano); (3) On-device AI (TinyML) — run inference on the sensor's microcontroller itself (ARM Cortex-M, <1MB RAM, <1mW power). Federated learning: instead of sending raw sensor data to a central server (privacy violation for home/health data), each device trains a local model on its own data and sends only model updates (gradients/weights) to a central server that aggregates them (FedAvg algorithm). The aggregated global model improves from all devices' data without ever seeing the data itself. ## Detailed Analysis TinyML pipeline: (1) Train large model (TF/PyTorch); (2) Quantize — reduce 32-bit float weights to 8-bit integers (post-training quantization, quantization-aware training); (3) Prune — remove unimportant weights; (4) Knowledge distillation — train a small "student" model to mimic a large "teacher"; (5) Convert to microcontroller format (TensorFlow Lite Micro, ONNX Runtime). Result: 50-500KB models running at 5-50ms inference time. ScienceDirect 2025 FL+TinyML survey: combining FL (privacy) + TinyML (efficiency) enables intelligent IoT without privacy-efficiency tradeoffs. Nature Comm 2025 personalized IoT: federated meta-learning (Per-FedAvg, pFedMe) learns a good initialization that adapts quickly — Model-Agnostic Meta-Learning (MAML) applied in federated setting, enabling 5-shot personalization. MDPI 2025 FL-for-IoT survey documents 15+ FL frameworks for IoT (TensorFlow Federated, PySyft, Flower, FedML). Applications: (1) Smart home — learning occupancy patterns for energy optimization without sending home data to cloud; (2) Wearables — personalized health monitoring (arrhythmia detection, fall detection) adapting to individual baselines; (3) Industrial IoT — predictive maintenance across factory machines; (4) Agriculture — soil and crop monitoring sensors. Key challenges: heterogeneous devices (phone vs. microcontroller), communication efficiency (sending model updates over low-bandwidth IoT protocols like LoRaWAN), and Byzantine robustness (malicious devices poisoning the global model). ## Further Reading - TensorFlow Lite for Microcontrollers (TinyML) - Flower: A Friendly Federated Learning Framework - Edge Impulse: TinyML Development Platform