Autonomous Geriatric Care System

Autonomous Elder Care Emergency System.

By: Ir. Narendran Ramasenderan, Vladislav Cherskoy, Mohamed Maahy Shuhad, And Umar Saeed

Introduction

In an era characterized by a growing elderly population and the increasing integration of smart home technologies into our daily lives, addressing the unique challenges faced by seniors in emergency situations has become a paramount concern. One of the most prevalent issues is the risk of falls among older adults, with global statistics indicating that one in four individuals aged 65 and above experiences a fall each year. Recognizing the urgency of providing timely assistance in such scenarios, a camera-based system driven by artificial intelligence, WiFi, and Bluetooth functionalities emerges as a promising solution. This innovative technology has the potential to autonomously detect emergency situations, such as a sudden fall, and promptly alert designated contacts and emergency services.

This proposed smart camera system embodies a proactive approach to elder care, utilizing cutting-edge technology to enhance the safety and well-being of older adults. Beyond the identification of emergency incidents, this intelligent system seamlessly integrates with other smart home devices, creating a comprehensive response to unforeseen circumstances. Through the utilization of features such as audio systems and smart locks, the system not only notifies designated contacts but can also engage neighbors and facilitate access for emergency personnel. By positioning the smart camera as a central component within the Matter Hub, it establishes a robust and interconnected network aimed at providing swift assistance in critical moments.

As the aging population continues to expand, the imperative for efficient and reliable elder care solutions becomes increasingly evident. This camera-based system, with its capacity to automatically detect emergency situations and trigger a rapid response, stands as a pioneering step towards addressing the unique needs of seniors in a technologically advanced age. Through the fusion of artificial intelligence, WiFi, and Bluetooth capabilities within the smart home ecosystem, this innovative approach promises a safer and more responsive environment for older individuals, ensuring their well-being is prioritized in times of need. In the subsequent sections, we delve into the intricacies of this solution, exploring its functionalities and the real-world impact it can have on the lives of older adults, offering both independence and peace of mind to this demographic and their families.

Our Concept

Building upon the conceptual framework outlined earlier, our proposed solution involves the implementation of an Internet of Things (IoT) based singular device designed to leverage advanced artificial intelligence training models, such as YOLOv8, for the automatic detection of emergencies involving the elderly. YOLOv8, renowned for its accuracy and efficiency, can be tailored through training to impeccably recognize specific occurrences based on nuanced body movements and visual cues. Strategically positioned at the top corners of the elderly person's residence, this singular device ensures constant monitoring through an integrated camera, capturing real-time data crucial for timely emergency response.

Our primary emphasis lies in the detection infrastructure, where the Jetson Orin Nano plays a pivotal role in our testing phase. This compact yet powerful microcomputer, equipped with a GPU, provides native support for running sophisticated visual detection programs. The USB port of the Jetson Orin Nano accommodates the attachment of a high-resolution camera, enabling detailed visual analysis. Facilitating wireless connectivity is the nRF7002 DK, serving as the interface for efficient communication. The choice of the nRF7002 DK is driven by its robust networking speeds, essential not only for expeditious communication with emergency services but also for seamless operation within the dynamic environment of a smart home.

By intricately integrating these cutting-edge technologies, our solution aims not only to detect emergencies promptly but also to communicate critical information to emergency services and designated contacts. The utilization of YOLOv8, Jetson Orin Nano, and nRF7002 DK ensures a comprehensive and technologically advanced approach to addressing the specific needs of the elderly in emergency situations within the context of a modern smart home. In the subsequent sections, we delve into the technical aspects of each component, exploring their functionalities and synergies to provide a detailed understanding of our proposed solution.

Testing

In the testing phase, we achieved significant milestones in validating the effectiveness of our proposed solution. YOLOv8, a key component of our detection infrastructure, was successfully trained to accurately identify instances where an elderly person requires emergency services. Our testing protocol involved utilizing both preexisting footage and generating our own recreations to compile a comprehensive dataset, ensuring the model's robust performance across diverse scenarios. When executed, the code demonstrated a high level of accuracy in promptly detecting emergency situations.

However, a noteworthy challenge surfaced during testing related to the compatibility of the nRF software with Linux on ARM architecture. The software exhibited compatibility issues, functioning solely on x86 architecture. While this presented a temporary hiccup, it is essential to highlight that such challenges are not insurmountable. Solutions range from potential updates to the software to alternative deployment methods, such as utilizing Docker or running a virtual machine. This hurdle does not undermine the overall success of the concept; rather, it points to areas for refinement and optimization in future iterations.

The testing phase, despite encountering a compatibility challenge, substantiates the feasibility and potential effectiveness of our proposed solution. With further development and fine-tuning, our concept holds promise in significantly enhancing the safety of elderly individuals by providing timely and accurate responses to emergency situations. In the subsequent sections, we delve into the specific details of our testing process, shedding light on the intricacies of training YOLOv8 and the challenges encountered during the integration of the nRF software.

Discussion

Our proposed autonomous elder care emergency system, integrating a camera-based solution and advanced technologies like YOLOv8, demonstrated promising outcomes during testing. The successful training of YOLOv8 to accurately detect emergency situations, particularly instances requiring immediate assistance for the elderly, underscores the robustness of our detection infrastructure. Utilizing both preexisting footage and recreated scenarios in our testing protocol contributed to a comprehensive dataset, ensuring adaptability across diverse situations. Despite encountering a significant challenge regarding the compatibility of the nRF software with Linux on ARM architecture, potential solutions such as software updates or alternative deployment methods offer avenues for addressing this hurdle in future iterations. The integration of YOLOv8, Jetson Orin Nano, and nRF7002 DK highlights the intricacies involved in creating a seamless and efficient detection infrastructure. The discussions around technical aspects provide insights into the synergy of these components, and the identification of areas for refinement during testing signifies a commitment to continual improvement, ensuring the system evolves to meet the dynamic demands of real-world scenarios.

Conclusion

In conclusion, our autonomous elder care emergency system represents a significant advancement in addressing the unique challenges faced by seniors in emergency situations within the context of smart home technologies. The camera-based system, driven by artificial intelligence, WiFi, and Bluetooth functionalities, offers a proactive and comprehensive approach to enhancing the safety and well-being of older adults. Despite encountering a compatibility challenge during the testing phase, the success of our detection infrastructure, particularly the training of YOLOv8, demonstrates the potential of the proposed solution. With a commitment to addressing challenges, refining components, and continuous development, our concept holds promise in significantly improving the timely and accurate response to emergency situations involving the elderly. This innovative approach aligns with the evolving landscape of smart home technologies, offering independence and peace of mind to the elderly and their families in an increasingly connected world.