Robotic assembly sequence generation requires many assembly predicates to generate optimal assembly sequences. Achieving such optimal assembly sequences is difficult as it requires huge search space. In the past literature, most of the researchers try to reduce the search space of assembly sequence planning problem by applying different methods such as computer-aided design-based methods, knowledge-based methods and artificial intelligence-based methods in terms of execution time and number of iterations by considering secondary parts (nuts, bolts, washers, fits, etc.). In this paper, stability graph concept has been introduced to represent the secondary part information for generating optimal assembly sequences. A comparison has been made without fruit fly in terms of execution time and number of iterations while generating optimal assembly sequences with and without consideration of secondary parts as primary parts. To check the effectiveness of the proposed method, three different industrial products with eight parts, 19 parts and 22 parts are considered for generating optimal assembly sequences. © 2019, King Fahd University of Petroleum & Minerals.