Disassembly facilitates the application of material and part recycling processes on product artifacts at the end of their useful lives. the artifact configuration greatly affects the disassembly efficiency. However, due to its complex nature, identifying the redesign modifications to improve the disassemblability of the configuration poses a major obstacle. a bottom-up approach using evolutionary computation principles is proposed to address this problem. an analogy to symbiotically related organisms is used, where one organism generates variant configurations and the other evaluates them by determining their associated disassembly sequences. These organisms evolve to obtain configurations that have superior disassembly efficiencies as compared to the original design. the efficiency measure used is obtained from the Configuration-Value model that we developed in an earlier work. an example is used to demonstrate the efficacy of the proposed method.