Assembly is one of the manufacturing processes, which occupies approximately 20% of the manufacturing cost of the product. In order to reduce the cost of the assembly, efficient assembly sequence is required. As the Assembly Sequence Planning (ASP) problem is the NP-hard problem, achieving the optimal assembly sequence for the complex products involves large amount of computational time and disk space. Moreover to achieve the optimal assembly sequence, assembly predicates (liaison data, stability data, and mechanical feasibility data) are to be satisfied by the assembly sequence. Extracting the assembly predicates involves huge disk space especially for the complex products. To reduce the effect of difficulty in achieving the optimal assembly sequence, in this paper an attempt is made to apply Design For Assembly (DFA) concept using Improved Firefly Algorithm (IFA). In this, initially, DFA predicates (contact data, material property data, functionality data, and relative motion data) are been extracted from the given assembly and are used to reduce the number of parts in the assembly, by which the number of levels of the assembly are been reduced. As the number of levels of the assembly is reduced, the energy effort to assemble the parts has been reduced. Later, IFA is used to obtain the optimal assembly sequence for the reduced levels of the product. The proposed methodology is implemented on machine block assembly and the results are compared with the general optimal assembly sequence planning techniques. © 2019, Springer Nature Singapore Pte Ltd.