Purpose The need to manage and monitor energy usage in companies and houses is becoming less than a luxury and more of a necessity, mostly due to financial reasons, but also due to the effects it has on climate change, being one of the main goals for sustainable development worldwide. This project aims to bring a system that’s capable of powering the user with information about the power usage of every appliance in an environment, along with the ability to remotely turn them on or off. For this, three boards were made in order to attend for a wide range of scenarios, be it a regular house with low power, single-phase appliances, to a corporate device that draws up to 100A on three-phase wiring. An artificial intelligence system was also developed to provide recommendations in the form of schedules that could potentially reduce the overall electricity waste without sacrificing comfort. The initial prototype was also installed inside in a laboratory so that we could validate its functionality while reading and controlling a high load, and also whilst only monitoring the general electricity meter. The proposed system seeks to cost only a fraction of the price of other products on the market, while also being easily scalable and practical enough to be used in a wide range of scenarios.

On the server side, a time series database servers as the storage for our collected data, serving as the main pillar for both data visualization and as a dataset for our learning techniques. To access it, a client facing website was set up, where the person responsible for the site’s administration can see current electricity usage, estimations, control the devices, schedule their operations and even opt for a recommended schedule made by our intelligent system. One basic application of this system is the power to audit the power consumption of a whole building, allowing the managers to know the main offenders when it comes to power draw, along with irregular usages, such as broken devices, decreasing financial costs related to electricity. On the other hand, instead of having to manually turn each device on or off by moving physically close to it, it is now possible to do so at the comfort of one’s own desk, by simply accessing the system through a computer or smartphone, increasing the quality of life. With the collected data, it is possible to infer patterns on the usage of not only devices, but also their surroundings, creating an adaptable, dynamic environment that’s able to optimize the appliances’ schedules for both comfort and power savings. As times goes by and the user further interacts with the system and makes choices about his personal schedules, the more the Artificial Intelligence can learn and perform better results while trying to achieve for a better consumption option that without neglecting the user’s comfort.