“…In order to incorporate the different testing levels in the model, we assume a cumulative exposure model as has been introduced in Sedyakin (1966) (see also Bagdonavičius, 1978; Nelson & Meeker, 1978). In fact, we follow the ideas developed in simple step‐stress testing as developed in, for example, Balakrishnan, Kundu, Ng, and Kannan (2007), Balakrishnan, Xie, and Kundu (2009), Kateri, Kamps, and Balakrishnan (2009), Mitra, Ganguly, Samanta, and Kundu (2013) (for reviews, see Balakrishnan 2009; Gouno & Balakrishnan, 2001; Kundu & Ganguly, 2017) where a similar idea has been successfully utilized. In our setting, we consider distribution functions F 0 , F 1 , … , F k − 1 corresponding to stages s 0 , s 1 , … , s k − 1 .…”