Machine Learning Algorithms and Sensor Technologies for Aging-In-Place Applications: A Review

The integration of artificial intelligence (AI) with wireless sensor networks (WSNs) has substantially advanced the capabilities of these systems to collect, process, and respond to large volumes of sensor data. Machine learning (ML) algorithms, a key component of AI, enable WSNs to detect patterns, trends, and anomalies, thereby enhancing their adaptability and performance in dynamic environments. One particularly impactful application is in aging-in-place systems, where elderly individuals can be continuously monitored for vital signs and daily activities, maintaining independence while benefiting from timely health and safety interventions. As the global elderly population grows, there is a pressing need for intelligent monitoring solutions to support independent living. This review aims to consolidate and critically evaluate existing ML algorithms and sensor technologies used in WSNs for aging-in-place applications. It introduces a structured taxonomy of ML approaches—including supervised, unsupervised, hybrid, and reinforcement learning (RL)—and presents a comparative analysis of wearable and environmental sensors. Unlike prior reviews, this work integrates technical performance evaluation with application-specific relevance, offering a unified view of algorithmic capabilities, sensor tradeoffs, and practical deployment challenges. Based on a systematic selection process, 124 peer-reviewed articles were analyzed. The review also identifies open research challenges, highlights limitations of current solutions, and proposes targeted future directions to guide the development of adaptive, reliable, and user-centered aging-in-place systems.