These days, there is a regularly expanding relocation of individuals to urban territories. Medicinal services administrations are a standout amongst the most testingviewpoints that is enormously influenced by the huge flood of individuals to downtown areas. Therefore, urban areas around the globe are putting vigorously in advanced change with an end goal to give more advantageous biological community to individuals. In such change, a great many homes are being outfitted with shrewd gadgets (e.g. brilliant meters, sensors and so on.) which produce monstrous volumes of fine-grained and indexical information that can be broke down to help keen city administrations. In this paper, we propose a model that uses brilliant home enormous information as methods for learning andfinding human movement designs for medicinal services applications. We propose the utilization of regular example mining, group investigation and forecast to quantify and dissect vitality use changes started by tenants' conduct. Since individuals' propensities are for the most part distinguished by ordinary schedules, finding these schedules enables us to perceive bizarre exercises that may demonstrate individuals' troubles in taking administer to themselves, for example, not planning sustenance or not utilizing shower/shower. Our places of business the need to dissect transient vitality utilization designs at the machine level, which is straightforwardly identified with human exercises. The information from keen meters is recursively mined in the quantum/information cut of 24 hours, and the outcomes are kept up crosswise over progressive mining works out.