The aim of this proposal is to develop algorithms for enabling mobile manipulators to accomplish purpose-driven objectives in complex indoor environments, specifically for indoor environment reconfiguration. This proposal aims to achieve the objective by i) developing mapping and knowledge representation methods tailored for dynamic indoor environments, ii) developing AI-guided optimal deployment of mobile manipulators for influencing changes to an indoor scene and iii) develop force-based manipulator interactions with movable objects. Using the framework, a team of mobile manipulators will demonstrate safe interaction and meaningful reconfiguration of a dynamic indoor environment.