Advanced Non-Intrusive Load Monitoring (NILM) research is important to help reduce energy consumption. Very-low-frequency approaches have traditionally faced challenges in separating appliance uses due to low discriminative information. The richer signatures available in high-frequency electrical data include many harmonic orders that have the potential to advance disaggregation. This has been explored to some extent, but not comprehensively due to a lack of an appropriate public dataset. This paper presents the development of a cost-effective energy monitoring system scalable for multiple entries while producing detailed measurements. We will detail our approach to creating a NILM dataset comprising both aggregate loads and individual appliance measurements, all while ensuring that the dataset is reproducible and accessible. Ultimately, the dataset can be used to validate NILM, and we show through the use of machine learning techniques that high-frequency features improve disaggregation accuracy when compared with traditional methods. This work addresses a critical gap in NILM research by detailing the design and implementation of a data acquisition system capable of generating rich and structured datasets that support precise energy consumption analysis and prepare the essential materials for advanced, real-time energy disaggregation and smart energy management applications.

Cetacean monitoring is key to their protection. Understanding their behavior relies on multi-channel and high-sampling-rate underwater acoustic recordings for identifying and tracking them in a passive way. However, a lot of energy and data storage is required, requiring frequent human maintenance operations. To cope with these constraints, an ultra-low power mixed-signal always-on wake-up is proposed. Based on pulse-pattern analysis, it can be used for triggering a multi-channel high-performance recorder only when cetacean clicks are detected, thus increasing autonomy and saving storage space. This detector is implemented as a mixed architecture making the most of analog and digital primitives: this combination drastically improves power consumption by processing high-frequency data using analog features and lower-frequency ones in a digital way. Furthermore, a bioacoustic expert system is proposed for improving detection accuracy (in ultra-low-power) via state machines. Power consumption of the system is lower than 30 μW in always-on mode, allowing an autonomy of 2 years on a single CR2032 battery cell with a high detection accuracy. The receiver operating characteristic (ROC) curve obtained has an area under curve of $85\%$ using expert rules and $75\%$ without it. This implementation provides an excellent trade-off between detection accuracy and power consumption. Focused on sperm whales, it can be tuned to detect other species emitting pulse trains. This approach facilitates biodiversity studies, reducing maintenance operations and allowing the use of lighter, more compact and portable recording equipment, as large batteries are no longer required. Additionally, recording only useful data helps to reduce the dataset labeling time.