• 

  Idea thumbnail

Inspiration

We were inspired to build EnergyGlance by the challenge many people face in understanding and predicting their energy consumption. High electricity bills and unexpected usage spikes are common frustrations. We wanted a solution that makes energy planning easy and transparent—something anyone could use, regardless of technical background.

What it does

EnergyGlance predicts tomorrow’s electricity usage using recent daily data. It calculates the average (mean) of the last three days, then uses it to forecast the next day’s usage. This helps users budget and plan, turning raw data into useful, actionable insights.

How we built it

Collected open energy usage data ( kWh/day kWh/day).

Cleaned the data by handling missing values and outliers with Python’s pandas library.

Forecasted next day’s consumption using the past three days:

Prediction n +

1

Usage n + Usage n − 1 + Usage n − 2 3 Prediction n+1 = 3 Usage n +Usage n−1 +Usage n−2

Evaluated performance with Mean Absolute Error (MAE):

MAE

1 N ∑

i

1 N ∣ Actual i − Predicted i ∣ MAE= N 1

i=1 ∑ N ∣Actual i −Predicted i ∣ Visualized results with matplotlib, showing actual vs predicted values.

Challenges we ran into

Finding reliable, complete energy usage datasets.

Handling inconsistent or missing values in real-world data.

Creating a balance between technical accuracy and user-friendly presentation.

Accomplishments that we're proud of

Built a working predictor that’s easy for anyone to use and understand.

Demonstrated the power of simple statistical methods for real-world forecasting.

Made the tool flexible for other time series data, not just energy usage.

What we learned

Simpler models can be very effective with clean data.

Data cleaning and visualization are as important as predictions.

Clear communication of results is crucial, especially for non-technical users.

What's next for EnergyGlance

Adding options for users to choose different averaging windows or forecasting methods.

Including additional features (like weather or time of week) to improve accuracy.

Building a simple web or mobile app interface for easy daily use.

Built With

• exel
• python