Intelligent Systems Lab

This project aims to develop an Energy Controlling Voice Enabled Intelligent Smart Home Ecosystem (ECO NISHE). The goal of ECO NISHE is to control basic appliances and monitor power consumption of a house with a user interface, based on vocal commands. The two basic concepts of the system are: a) simple setup, based on non-intrusive methods and b) straightforward usage of the system, even from individuals not familiar with technology or people with physical impairments, such as restricted mobility and/or vision.

ECO NISHE is a system integrating hardware and software and is focused on the control of electrical devices with dual purpose:

• improvement of residents’ quality of life, including elderly ones and people with mobility and/or vision problems and
• constant update and awareness regarding electrical power consumption and how it is distributed among existing heavy load appliances, resulting in a more efficient energy profile of the house.

ECO NISHE consists of the following components:

1. a tableware central control unit,
2. a power sampling device, measuring the aggregate electricity consumption which is usually placed at the electrical panel,
3. interior and exterior wireless multisensory system (sensors of temperature, humidity, luminosity, smoke and CO2),
4. smart thermostats, plus and switches which replace the respective traditional ones that already exist in the house,
5. intelligent software system using Machine Learning technology, which disaggregates the total energy consumption into the basic heavy load appliances, without the need of replacing them with smart ones or adding any extra equipment,
6. intelligent software with voice recognition and speaking capabilities, enabling the users to communicate with the house in their natural language,
7. middleware and communication protocols for the coordination between hardware components and services of the whole system and
8. software application recording anonymous statistical data about energy consumption in each house having ECO NISHE installed and distributing this information as Open Linked Data, for research purposes and playing a primary role in the overall energy policy of the state.

The main functionalities of the system include:

1. updating the residents via voice messages and other applications (web and mobile apps) regarding the following:
   1. overall energy profile of the house (consumption per day/ month/ year, time periods with maximum/ minimum consumption)
   2. energy comparison and classification relative to other energy equivalent houses of the same area (only after owner’s permission to share relative data)
   3. real time electricity consumption of specific types of appliances (e.g. lighting, heating/ cooling etc.)
   4. distribution of daily/ monthly/ annual electricity consumption in categories (e.g. lighting, heating/ cooling etc.)
   5. current indoor and outdoor environmental conditions (temperature, humidity etc.)
2. control and monitor through voice commands and applications of the following:
   1. the desired temperature in the house
   2. the condition of appliances which are connected with smart plugs and switches
3. advising residents on how to save energy and making them aware of the overall energy profile of the house (e.g. recommending level settings of the thermostat regarding the temperature (indoors or outdoors), the hour of the day, outside luminosity and residents’ behavioral patterns.
4. presenting warnings to the residents for events requiring attention such as:
   1. sudden rise in temperature of the house
   2. abnormal CO2 levels
   3. irregular humidity levels
   4. uncommon increase of electrical consumption.