Low-Cost LIDAR-Based Monitoring System for Fall Detection

Every year, over 30% of individuals aged 65 and above experience fall, leading to potential physical and psychological harm. This is particularly concerning for those who live independently and lack immediate assistance. To address this issue, numerous studies in the field have focused on early fall detection of the elderly, employing diverse sensors and algorithms. In this paper, we present a low-cost fall detection system based on Laser Imaging, Detection And Ranging (LIDAR) technology, specifically designed for older individuals residing alone, monitoring daily routines without disruption and without camera, to respect user privacy, as these are essential factors for user acceptance and overall effectiveness. The system's significance lies in its capacity to function with minimal computational resources and its high level of interpretability. This is due to our solution being based on Finite State Machines (FSM), which contain clear rules, distinguishing between different states such as no human presence, human presence, and a fall, offering a more transparent and interpretable detection process. These characteristics contrast with traditional methods such as Support Vector Machine (SVM) and Artificial Neural Network (ANN), which are more computationally expensive and operate as black boxes. This approach combines simplicity, low computational load, privacy preservation, low cost, interpretability, and high accuracy. It represents an advancement in enhancing the care and safety of older adults living independently.