Comparing environmental sustainability in the Smart Home between Open Source and Vendor Systems

The expansion of smart home technology has led to increasing concerns about environmental sustainability, particularly regarding CO₂ emissions from continuously operating smart home management systems (SHMS). While existing research has examined individual smart home components, a significant gap exists in understanding the carbon footprint of the central management systems that coordinate these ecosystems. This knowledge gap is crucial as these systems operate 24/7 and their environmental impact directly influences the overall sustainability of smart home adoption. This thesis addresses the problem of quantifying and comparing CO₂ emissions between vendor-based and open source smart home management systems. The primary objective is to develop a methodology for estimating both embodied and operational emissions, enabling evidence-based comparisons between different technological approaches to home automation. The approach involved establishing a controlled experimental environment using the Internet Microscope methodology for traffic monitoring and inline power meters for energy consumption measurement. Representative systems including Philips Hue Bridge and Amazon Echo Pop (vendor-based) versus Home Assistant Green and Voice PE (open source) were analyzed, examining idle consumption, voice control interactions, and software updates. Embodied emissions were estimated using an established weight-based methodology, while operational emissions incorporated both energy consumption and data traffic using the Sustainable Web Design model. The study's experiments demonstrate that SHMS consume between 1.5-1.9W during idle operation, with annual emissions ranging from 3.65-7.25 kg CO₂e per device. The open source voice assistant achieved 33% lower energy consumption and eliminated cloud-dependent data transmission compared to the vendor alternative, though with increased response times. Results indicate that operational emissions dominate over embodied emissions in long-term usage scenarios, with devices exceeding their manufacturing carbon footprint through idle consumption alone within approximately two years. These findings suggest that power efficiency should be prioritized in SHMS design and selection criteria. This work contributes to the field by providing the first systematic comparison of CO₂ emissions between vendor and open source smart home management approaches, establishing a foundation for environmentally informed decision-making in smart home technology adoption.