With the rapidly aging population, there is a need to detect elderly's activity, monitor their health, alert health care personnel, and provide real-time and long-term access to the generated sensor data. With the advances in sensor technologies, communication protocols, computing power, and cloud and edge services, it is now possible to build smart assistive living systems to improve people's lives. In this research, we present a Context-aware and private real-time reporting aging in place system. The proposed system, we call it smart carpet, consists of a sensor pad placed under a carpet; the electronics reads walking activity to provide an automated health monitoring and alert system. We extended the system's functionalities to improve its ability to detect falls, measure gait, and count the number of people traversing the carpet (socializing). In an urgent and time-sensitive situation, we need to provide a real-time notification. We propose a cooperative cloudlet model, where the sensors' data will be sent to the nearest Cloudlet for analysis and extracting real-time decisions in minimal delay. Results showed that our system could assist the elderly in detecting falls with 95% sensitivity and 85% specificity. Measuring and estimating their gait with a mean percentage error difference to GAITRite 1.43% in walking speed; hence, predicting a fall risk and counting people's plurality socializing with the elderly with an average accuracy of 100%. We evaluated our system's improvements in a controlled environment. We are looking forward to deploying the system in a nursing home(after the COVID-19 is over) for more data gathering and validations.