This study aims to create an on-demand intelligent space using a detachable arm that can be installed only when needed. To enable stable task performance in various environments, this study proposes a method to co-optimize the arm's placement and mounting surface shape. This is achieved by solving a single nonlinear optimization problem that integrates four key requirements: geometric conditions, mechanical stability, design of the mounting surface shape, and the trade-off between stability and space-saving. To physically realize the diverse shapes derived from this optimization, a modular outrigger mechanism is designed. Experiments with a real robot demonstrated that the system adaptively generates the minimum necessary mounting surface for different task requirements, such as operating in confined spaces or withstanding large forces.
