I am analysing survival data of raptors. I evaluated several competing functions with aft models and found that Weibull functions outperform the others. The shape parameter is always bigger than 1, from where I derive that survival probability decreases with age (higher risk with time). The best model is one considering several different subgroups. I am trying to estimate age of onset of senescence for each group, but I don't know exactly how to extract it from the models (or from the data). Can anyone help?