Early warning systems (EWS) are increasingly used for slope failure risk management. The most crucial tasks of EWS are prediction of the slope failure time and establishing alert procedures, however implementation of these two tasks is challenging due to the impact of various factors. This study aims to investigate both of these tasks using dimensionless processing methods in order to establish a general framework for EWS. A historical slope failure database is firstly established containing 50 failure cases and 50 non-failure cases to study our methods. A dimensionless inverse velocity method (DINV) is proposed to provide a general solution framework for calculating the time of failure. Additionally, linear and non-linear DINV trends are used to establish a failure time window to evaluate the potential failure time frame. For the alert procedure, two proposed dimensionless indicators are used to define a series of threshold levels. These two indicators can be combined as a potential tool for determining whether a slope with acceleration will eventually fail. The system adopts dimensionless processing methods and considers different slope failures with similar acceleration patterns. Therefore, this approach can be used as a general framework to assess the slope failure risk.