The planning of an assembly sequence is an fundamental task in designing a robotic assembly system because it determines the system lay-out and its efficiency. Therefore, this paper presents a method for the automatic generation of robotic assembly sequences based on the verification of a disassemblability for a part to be diassembled. To derive the disassemblability of a candidate part, first we inference collision free assembly directions by extracting separable directions for the part, and calculate the separability which gives informations as to how the part can be easily separated. Using the result, we determine the disassemblability defined by the separability and stability cost. The separability cost represents a facility of the part disassembly operation, and the stability cost represents a degree of the stability for the base assembly motion which gives a design guidance of jig/fixture. The proposed method provides a unique solution of finding a cost effective assembly plan, and the feasibility of calculating robot motion programs by evaluating separable directions in flexible manufacturing application. In determining the system lay-out, however, one more important factor to be considered is the line balancing problem. This is because, when planning assembly lay-out, it requires a sophisticated planning method. For the efficient inference of line balanced assembly sequences, many works have been reported on how to evaluate an assembly work time at each work station. However, most of them have some limitations in that they use cumbersome user query or approximated time data without considering assembly conditions. To overcome such criticism, this paper proposes a new approach to mathematically verify assembly conditions based on disassemblability. Based upon the results, we present a method of evaluating assembly work time using neural networks. The proposed method provides an effective means of solving the line balancing problem and gives a design guidance of planning assembly lay-out in flexible assembly application. Example studies are given to illustrate the concepts and procedure of the proposed schemes.