DevilSight

Inspiration

What it does

Software that analyzes patterns of electricity consumption by area, is capable of determining what your electricity consumption will be in the future given the patterns of past years and being able to see which areas consume electricity in an atypical way in order to detect possible electricity theft, so that the provider can be aware of the situation and check it out as soon as possible.

How we built it

Our team was focused on 2 main parts, the development of the web application using HTML, CSS and Django, while the other half was focused on the development of a Machine Learning model that could efficiently predict consumption patterns by zone.

Challenges we ran into

A large part of the difficulty that the prototype had was finding a dataset that contains data on the consumption of electrical power in different zones.

Another difficulty that we presented was the training of the model itself, since we noticed that with the passage of time while the model was being trained, these instead of decreasing the model error seemed to increase it, we fixed this with Early Stopping.

Accomplishments that we're proud of

Train a functional machine learning model that meets our expectations and a user-friendly web application that meets our needs

What we learned

Working under pressure with a team. It is important everybody knows what they have to do, so everything is coming on time and in the way is planned.

What's next for DevilSight

The way this system works does it capable to use It almost everywhere. It can help every electric provider optimizing times, human capital and lowering the amount of wasted energy, translated on bigger incomes.