Multi-modeling and co-simulation are one of the solutions for dealing with complex systems. In this paper, we propose to apply the AA4MM framework to the co-simulation of a smart space heating "complex" environment with 2 objectives. Our first contribution is the development of AA4MM-FMI (Agent and Artifact for Multiple Models coordination for FMI). AA4MM is a modeling and simulation framework that can be used to implement the "master" in a multiple FMU (Functional Mock-up Unit) co-simulation of the FMI (Functional Mock-up Interface) standard with a fully decentralized view and intrinsic coordination. Our second contribution is to apply and evaluate this solution by creating a hierarchy of smart space models that are exploiting house geometry information. Each individual room is represented as an independent FMU taking inputs of the room's geometric information (surface area, volume, insulation, etc.), target temperature, and neighboring rooms. The neighboring rooms are connected to each other using the AA4MM-FMI framework. This project aims to couple an electrical heating-based simulation with networking event-based simulations to gain intuition for how house geometry affects efficient heating and network connectivity. Using this smart space house heating problem, we are able to test AA4MM-FMI (our novel "master") for FMI and multi-simulation.