Siirry sisältöön
AI-based energy management in SMEs

It is estimated that approximately 13 % of the global total energy demand is used in small and medium organizations (SMEs). As energy efficiency is a major objective of the European Green Deal and considering SMEs’ massive energy consumption and environmental impact, the solutions of energy management in SMEs must be prioritized.


Ari Alamäki

yliopettaja, myynnin kehittäminen ja digitalisaatio
Haaga-Helia ammattikorkeakoulu

Published : 10.10.2022

It is estimated that the share of SME energy consumption on gross inland consumption ranges from 9 % to 18 % (Leap 4 SME 2021). This requires serious attention by the EU governments which are striving for the efficient use of resources and to curb climate change. The Energy Efficiency Directive (EED) urges SMEs to undertake energy audits and implement stringent policies for substantial energy reduction.

The prime motivation to pursue energy reduction in SMEs is not only to alleviate the cost but also to combat climate change. Sustainable energy use in SMEs, including the application of renewable energy sources and improved energy efficiency performance, can not only improve the environment and fight climate change but can save companies money and thus boost their competitiveness. 

That said, energy management in SMEs is a hugely neglected area due to a lack of awareness and motivation, little to no expertise, financial constraints, and the tendency to prioritize other investments over energy consumption.

Most SMEs do not even have a concrete energy monitoring system, and they rely on energy bills and meters to estimate their energy consumption (Southernwood et al. 2021). Quite the contrary, an aggressive energy reduction entails acquiring and analyzing data on consumption patterns, usage history, energy consumption modes of the devices and their workload, user preferences, hours of maximum and minimum activity, security and safety emergencies, and energy-hungry sources.

Since it is not trivial to analyze all these factors manually and implement an energy reduction policy on the fly, the process requires an intelligent, automated mechanism that can optimize the energy consumption of an SME.

Opportunities of AI for reducing energy consumption

While AI is gaining increasing adoption in SMEs for the automation of routine tasks, it can also offer an opportunity to alter the way an SME manages and consumes energy (Vinuesa et al. 2020).

AI can leverage deep learning and machine learning algorithms to gain insights into the energy operations of an SME and select an optimal, cost-efficient, and environment-friendly energy management policy in real-time. AI can be applied to develop an IoT-based system that collects energy consumption data under different conditions, times, and workloads to analyze and optimize energy usage in SMEs.

An AI system comprising different AI agents can be deployed to gather energy consumption data from different energy consumption sources to find the underlying energy consumption patterns. Thus, learning an efficient energy management policy for it, and minimizing the overall energy management in an SME.

These policies include selectively turning on/off the energy sources and/or switching their energy consumption modes based on several factors such as usage history, amount of energy consumption, the time required to transit from one energy state to another, environmental conditions, source’s criticality, and demand-response requirements, etc.

Heating, ventilation and airconditioning optimized by AI

The energy consumption of an HVAC (Heating, Ventilation, and Airconditioning) system in an SME can be controlled by an AI agent for optimal decision-making regarding energy management based on collected data of energy consumption history, outside temperature, ambient temperature, gathering frequency of individuals in a particular area, humidity, hour of the day, to name a few.

Such a system will not only result in energy savings but will also provide a proactive approach for maintaining a steady indoor temperature in an SME. Apart from that, the dynamic energy management of an HVAC system is equally instrumental in the SMEs that maintain their data center or provide data center services.

The energy consumption of HVAC systems in data centers accounts for 30-40 % of the overall energy consumption (Yang et al. 2021). By collecting and analyzing the data related to workload patterns, level of quality of service, computational requirements, and other aforementioned environmental parameters, an AI system can learn an energy management policy to significantly reduce the overall energy consumption of a data center in an SME.

Awareness for energy savings in SMEs

Before taking advantage of the immense potential of AI for energy reduction in SMEs, it is particularly important to first raise awareness about the importance of energy reduction and its benefits in SMEs.

This could be done by large-scale training, awareness seminars, and special courses. Now is the time to define and recognize opportunities and design AI-enhanced solutions to monitor, manage and control automatically energy consumption in SMEs.


Leap 4 SME, Energy Audit Policies to Drive Energy Efficiency, online report, April 2021

Southernwood, Joanna, et al. 2021. Energy Efficiency Solutions for Small and Medium-Sized Enterprises. Multidisciplinary Digital Publishing Institute Proceedings, 65.1: 19.

Vinuesa, R. & et al. 2020. The role of artificial intelligence in achieving the Sustainable Development Goals. Nature Communications 11(233).

Yang, Z., Du, J., Lin, Y., Du, Z., Xia, L., Zhao, Q., & Guan, X. 2021. Increasing the energy efficiency of a data center based on machine learning. Journal of Industrial Ecology, 26(1), 323-335.