Gupta, ManuMary, ShrutiTrivedi, JayGoyal, JaiMaan, PriyankaSingh, MagendraWith the increasing number of applications and usage of mobile phones, the limited storage of battery needs to be managed properly so that the battery does not drain when it is desperately required by the user. It is important for the user to know how long her/his battery will last. There has been a significant study in the area of mobile phone battery lifetime estimation. Some researchers have studied user patterns from device logs, but only few of them have considered charging cycles into account for study. This paper proposes a new on-device approach to estimate battery lifetime based on user history, using both charging and discharging cycles. The charging cycle is processed as well to estimate how much battery has been consumed in that cycle. The proposed algorithm fetches the battery discharge governing features from user handset, processes it, and creates the history for the user. Now based on the user’s individual history, the battery lifetime is estimated. The dataset for user history has been collected from October 2019 to December 2019. We have been successfully able to estimate battery lifetime using the proposed algorithm with a root mean square difference between estimated battery lifetime and the time battery actually lasted to be 160 min for any typical user, which is better by an average of 196 min when compared to results of the existing battery lifetime estimator present in the smartphones of the participant users.