While there is an increasing integration of connected devices in buildings to reduce energy consumption, little evidence exists regarding the effective environmental impact of the smart devices installed. One goal of this study is to comparatively assess the impact of integrating connected devices on the environmental performances of existing buildings. Additionally, it aims to discuss the added value regarding environmental performance between a strategy of integrating connected devices and a refurbishment strategy. A non-refurbished connected building will thus be compared with a non-connected refurbished building. The approach's novelty lies in carrying out an uncertainty analysis, which addresses uncertainties in the environmental modelling of sensors, heating gains, and other influential building parameters, such as occupancy scenarios and building lifespan. The Life Cycle Assessment model is built on the results of a dynamic building energy simulation and integrates environmental data on connected devices from ecoinvent. Environmental indicators are calculated using ImpactWorld+. The methodology is applied to a multi-storey residential tower in France, characterised by a low insulation level. The alternatives compared include the original building without sensors, two connected alternatives (monitoring and management) that lead to a reduction of heating loads by 10 and 30% respectively, and a refurbished alternative. Results confidently show that using connected devices is always better than keeping the original building state. Yet, refurbishing the building is the preferable option across almost all indicators. Finally, the results point out that energy gains allowed by connected devices should be better quantified in further work through measurement campaigns.