At the end of November 2021, a regional energy conference was held in Opatija (Croatia), co-organized by the INSULAE partner Regional Energy Agency Kvarner and the Regional Development Agency of Primorje-Gorski Kotar County, dedicated to the green transition of Primorje-Gorski Kotar County and the financing possibilities through ESI Funds 2021-2027.

Mr. Damir Medved from Ericsson Nikola Tesla Ltd. presented the first results of the research conducted within the INSULAE pilot project on the Unije island. The research aimed to determine individual patterns of behavior of the island’s inhabitants by processing large datasets.

In the previous phases of the INSULAE project, measuring devices were installed in a dozen of households on the island, enabling the collection of relevant data and their subsequent analysis. The process has been carried out in 5 steps:

  • Step 1. Device analysis. Determining the household variant by mapping all devices related to their characteristics and behavior (e.g. washing machine characteristics and usage characteristics);
  • Step 2. Set goals (e.g. energy efficiency and cost reduction versus resilience);
  • Step 3. Habits analysis. Analysis of user needs concerning a specific use scenario (e.g. elderly people living alone on Unije, a young married couple with an active lifestyle who occasionally comes to Unije, a young family with two small children that wants to move to the island, etc.);
  • Step 4. Household consumption/production profile. Creating a reliable Unije household model and defining usage scenarios;
  • Step 5. Analysis of the costs and benefits of the proposed approach for islanders.

This approach allowed the creation of standardized patterns of behavior that can later be replicated in more complex models:

Figure 1: INSULAE – Household energy production/consumpion profiling

This typically represents an entrance into systems for simulating energy production/consumption, and then into automation systems that enable real-time energy flow management based on machine learning technologies and artificial intelligence.

In the case of Unije island, 3 models of behavior have been created that can be mapped to users with the following characteristics:

  1. A middle-aged family consisting of 4 people (parents and two children), with the age of the parents ranging from 30 to 50 years;
  2. An older couple, over 65 years of age;
  3. A young single traveler, whose age ranges from 24 to 40 years.

The models defined in this way are used to create more detailed models of energy consumption, which can significantly improve the process of managing the energy flows of the island:

Figure 2: INSULAE – Smart Island Platform user analysis dashboard (first released version for beta testers)

Given the experience gained, it can be concluded that the chosen methodology is correct. It has been recorded that energy consumption in a smart home/building can be achieved without major changes in the habits of household residents, but significant savings can only be achieved by a complete change of mindset.

In the next phase of the INSULAE project, it is planned to further expand the installed device database and collect a much larger amount of data, which should enable more relevant conclusions, i.e. the start of the final project phase during which the real-time consumer management concept will be demonstrated and proposed.

Figure 3 Event panelists

Figure 4 Damir Medved, Ericsson Nikola Tesla Ltd.

Software photos credentials:

Ericsson Nikola Tesla Ltd.

Event photos credentials:

Regional Energy Agency Kvarner