“…The SA by Rudek & Rudek (2012) cannot be compared to the ACO and CPSO developed by Huang & Yu (2016) , since Huang & Yu (2016) considers deteriorating maintenance duration while Rudek & Rudek (2012) focuses on deterioration of job processing times. Seif et al (2018) propose a GA-inspired algorithm which they name ALG and show that the ALG outperforms a standard GA. Aggoune (2003) , Kaabi, Varnier, & Zerhouni (2004) , Allaoui, Lamouri, Artiba, & E.Aghezzaf (2008) , Gara-Ali & Espinouse (2014 , Hadda (2015) , Mosheiov et al (2018) , Besbes, Teghem, & Loukil (2010) , Luo, Huang, Zhang, & Dai (2011) , Benbouzid-Si Tayeb, Guebli, Bessadi, Varnier, & Zerhouni (2011) , Nouri, Fattahi, & Ramezanian (2013 , Benbouzid-Si Tayeb & Belkaaloul (2014) , Boudjelida (2019) and Chansombat, Pongcharoen, & Hicks (2019) consider scheduling PM in a flexible window. Most studies use strict flexible windows, but in the studies by Kaabi et al (2004) , Benbouzid-Si Tayeb et al (2011) , Nouri et al (2013) , Benbouzid-Si Tayeb & Belkaaloul (2014 and Boudjelida (2019) , maintenance may be delayed or advanced beyond the time window, imposing earliness and tardiness penalties on the objective function.…”