The ability to handle and process data for information extraction is getting more and more important. Using extracted data from the business to improve productivity is seen as an important part in developing the business processes. In this thesis, industrial robots and their survival times are analyzed. The work is about predicting the probability that a specific robot will fail during a specified time period. Also, survival analysis is conducted where the median lifetime and conditional median lifetime for industrial robots are estimated. Two approaches are used, logistic regression and survival analysis. A logistic regression model is made to predict the probability for different industrial robots to break during a specified time period. The logistic model achieves an accuracy of 0.694 with even higher accuracy regarding high – and low risk robots. The survival analysis uses a Cox PH model to check validity for proportional hazards and then a parametric model with Weibull distribution is fitted. The parametrical survival model is used to estimate the median lifetime and the remaining median lifetime for the robots. The estimated probabilities and lifetimes can be used as an indication of which robots are in risk of failure.