Recently I have applied for a application project funding at FEEIT together with my colleague Zivko Kokolanski. Today, I am happy to announce that the funds for the project proposal titled “Home Energy Return of Investment” (HERoI) are granted.
Bellow you can find a short description of the scope of HERoI project.
The smart home is becoming a reality and there are many possible improvements one can consider from the available data and connectivity. There are two main beneficiaries in this situation:
- The power distribution companies may properly plan the demand on the small customers’ side, generating a personal user consumption/production profile. Sharing this profile with the distribution companies will significantly improve their planning for electrical power production and distribution.
- Home users may improve their efficiency and reduce the cost. Regular users don’t want to have to prompt their devices to complete an action. They want their devices to use data, analytics and sensors to work on their own. Therefore a smart home assistant system build around the smart home devices, designed particularly to anticipate the user’s needs is the next logical step for smart home development. The result of implementation of such an assistant system will be improved cost savings for the user and improved environmental impact from energy efficiency and monitoring.
Our goal is to design implement simple hardware and machine learning algorithms that will enable both the consumers and electrical power distribution companies to exploit the benefits from improved consumption profiles. Specifically the goals might be summed up as:
- Integration of machine learning algorithms with predictive control to improve the prediction of consumption/production in smart homes.
- Generation of consumption/production profiles of home users, taking into consideration smart home appliances and power generation/storage in the smart home.
- Detection of patterns in smart home and their automation with minimal user interaction.
We expect to design new algorithms that benefit from the integration of machine learning models with control algorithms. These algorithms will result in reduction of the energy consumption; reducing the carbon footprint; improved power production planning and; home user’s cost reduction.
The complete project application (in Macedonian) is available upon request.
More details to follow.