People with reduced or impaired mobility often rely on mechanical support. The current designs of such artificial support tend to focus on functional compensation, which does not necessarily encourage the user to walk independently nor help recover to a normal gait in the long run. To the contrary, some aids are known to enforce the user’s dependency on the support and as a result, fail to exploit human adaptability. For the design of adaptable human-machine interaction, there is a fundamental need for quantitative models that consider, not only musculo-skeletal patterns of gait, but also cognitive modes involved (e.g., willingness to walk on their own, or the degree to which the user depends on machine). To address the gap, this paper aims to develop and test an integrative cognitive-musculoskeletal model that can represent and simulate human gait for the humans with impaired mobility.