Entropy Platform

ENTROPY regards an energy-aware IT ecosystem aiming to support energy efficiency in the buildings sector through behavioral change of the occupants with regards to their daily energy consumption patterns. The main distinguishing characteristic of ENTROPY is that it exploits the advantages provided by a set of novel ICT technologies for enabling the design, development and provision of personalized energy management and awareness services in smart buildings. The philosophy of the proposed approach is based on the provision of personalized services that can lead to behavioral change through energy consumption awareness and motives provided to occupants based on their behavioral profile. The set of services are supported through the ENTROPY platform and are made available to mobile application developers for developing personalised applications and serious games.


The main adopted and integrated technologies in the ENTROPY platform include Internet of Things (IoT), information fusion, semantic web, rule-based recommendations, big data mining, and analysis mechanisms. Novel IoT node configuration, networking, and efficient data aggregation mechanisms, including mobile crowd-sensing mechanisms, are applied for the interconnection of any type of sensor (e.g., low-cost sensors such as Arduino and Raspberry Pi), a collection of data, and the application of data quality enhancement mechanisms (e.g., removal of outliers, fix missing values in time-series data). Information processing, semantic mapping, and fusion mechanisms are applied for representing the collected data in a unified way, boosting in this way their exploitability and interoperability with existing services, as well as their interlinking with available data. Recommendation mechanisms are applied for real-time reasoning over the available data and provision of suggestions for personalized actions that can lead to improving energy efficiency through behavioral change. Big data mining and analysis mechanisms are also supported for producing behavioral and energy consumption analytics, targeting at providing advanced insights and increasing the energy-awareness level of end users.


All the aforementioned technologies are supported through the ENTROPY Platform that comprises the basis for the consumption of existing services, as well as the design and development of further energy management and awareness services, personalized mobile applications, and serious games.


Two types of users are considered in ENTROPY, namely, campaign managers and end users. Campaign managers are responsible for setting up an energy efficiency campaign and may consist of smart buildings administrators, energy efficiency experts, data scientists and behavioral scientists. End users regard the set of users participating in the campaign and they may consist of citizens, students, academic personnel, employees in enterprises, etc. An indicative workflow followed for the realisation of an energy efficiency campaign is depicted at Figure 1.

Figure 1: Energy Efficiency Campaign Workflow


A high-level view of the ENTROPY energy-aware IT ecosystem architectural approach is provided at Figure 2. As depicted, a layered architecture is followed with discrete layers for IoT management and data aggregation, data representation and fusion, smart energy management services and end user applications. The IoT management and data aggregation layer is responsible for IoT nodes registration, management and data aggregation and cleaning functionalities at the edge part of the infrastructure. The data representation and fusion layer is responsible for representing the collected data based on a set of defined semantic models as well as supporting a set of data fusion mechanisms over active data streams. The smart energy management services layer is responsible for providing advanced analytics and recommendations to end users, as well as incorporating learning techniques for continuously exploiting the produced output by each service. The end user applications layer is responsible for the design of personalized mobile applications and web-based serious games able to take advantage of the set of services provided by the lower layers. Following this, detailed information is provided for the designed and implemented mechanisms per layer.

Figure 2: ENTROPY Platform Architectural Approach


The ENTROPY Platform is up and running in the three ENTROPY pilots, well interconnected with the developed personalised applications and serious game, and supporting a numerous set of end users and campaigns. Based on the continuous feedback received from the set of pilot campaigns, validation of the efficiency of the supported mechanisms along with updated and improvements -where required- is taking place.


A walkthrough in the ENTROPY Platform is available at https://www.youtube.com/watch?v=qKJrcwUG88w

legavEntropy Platform