Utforska implementeringen av AI-teknik inom ett litet och medelstort fastighetsförvaltningsföretag: En fallstudie

Sammanfattning

Abstract. The facility management (FM) industry is under pressure to reduce operational costs and environmental impact, yet small and medium-sized enterprises (SMEs) face significant barriers to adopting advanced technologies like artificial intelligence (AI), including limited resources, technical complexity, and organizational inertia. Despite growing interest in AI-driven energy optimization, there is a lack of empirical research on real-world implementation within SME contexts. This study addresses that gap by investigating the early-stage implementation of an AI-based heating optimization system in a Swedish SME FM company, inspired by Actor-Network Theory (ANT) to understand how both human and non-human actors interact to shape the outcome. A single-case study design was employed, combining semi-structured interviews with a facility manager and AI software supplier, quantitative energy consumption data, and technical documentation. The analysis focused on how real-time IoT sensor data and predictive algorithms were integrated into existing building systems to optimize heating, improve indoor climate, and reduce energy consumption. Findings show an initial reduction of 5.2 kWh/m² in heating energy use, which is below the supplier's estimated 12 kWh/m² but promising given the short evaluation period. This study reveals that the implementation is still in a formative phase, requiring stronger alignment between actors and a full year of operational data before drawing long-term conclusions. This study provides practical insight into the socio-technical dynamics of AI adoption in FM and highlights key conditions for successful scaling of AI in SME settings. Keywords: Artificial Intelligence (AI), Facility Mangement (FM), Energy efficiency