Tsetlin Machine enables real-time, low-memory NILM on MCUs

Researchers have developed a novel Non-Intrusive Load Monitoring (NILM) system utilizing a Tsetlin Machine (TM) framework, designed for real-time operation on resource-constrained Microcontroller Units (MCUs). This approach reformulates the energy consumption estimation problem as a classification task, enabling privacy-preserving edge deployment by processing sensitive household data locally. The TM-based system demonstrates strong performance, achieving 90% precision and 96% recall for classifying two appliances and 77% precision with 80% recall for four appliances on the REDD dataset. Notably, the trained model requires only 18 KB of flash memory and exhibits an inference latency of 0.43 ms on an ESP32, making it highly suitable for embedded NILM applications. AI

IMPACT Enables privacy-preserving, real-time energy monitoring on low-power edge devices.

RANK_REASON Academic paper detailing a new machine learning approach for a specific application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• ESP32
• Microcontroller Units
• Non-intrusive load monitoring approaches for disaggregated energy sensing: a survey
• REDD dataset
• Tsetlin Machine

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →