There has been a global increase in the aged population in recent years, so the number of older adults living alone at home also keeps increasing. Older adults often get in trouble or risk at home, it is quite hazardous to those who live alone. During COVID-19, older adults are vulnerable to viruses and more likely to pass away unnoticed to anyone. Taking advantage of smart techniques, this study aims to ensure indoor safety of older adults living alone based on digital information on indoor movements. Motion sensors should be placed to collect movement data at home with privacy protection. Based on the time series data of motion sensors under normal circumstances, the indoor movement trajectory of the elderly can be generated. Sequential pattern mining is used to identify older adults’ indoor movement patterns in daily life. The smart home provided by CASAS is adopted as the study case. Findings show movement patterns that occur frequently can be identified successfully by sequential pattern mining, and movement patterns are various in different periods of a day. The movement patterns are the basis of further identification of abnormal movements for alerting abnormal behaviors during specific periods, reporting near-miss events, and calling for first-aid services. It would be a convenient and innovative sensor-aided way to reduce the risks of older adults living alone and prevent them from unexpected injuries or deaths caused by missing the best timing of treatment.