Data Visualization and Interaction with OSISoft and UCDavis Energy Data


We wanted to help make UC Davis an even greener campus through data analysis through more intuitive visualizations. We also wanted to allow easy access to the data by making an Amazon Alexa skill for voice activated data access. Hopefully, predictive modeling would emphasize what areas to focus on for minimizing energy consumption.

What it does

  • Data Visualization
  • Predictive Modeling
  • Alexa Skill

How we built it

We scraped the data with HTTR and accessed those elements and searched through the API and mapped out the buildings by what utilities they have. We found our final goal in terms of data acquisition by getting a data frame that gets every unique combination by building it with utility type and variable type. Used R and Shiny for the visualization, and Javascript and Amazon Web Services for Alexa Skill.

Challenges we ran into

Cleaning Data Finding significant relationships between Attributes Data Abnormality skews Predictive Modeling Alexa Skill Formatting and Terminology

What we learned

Coming up with general strategies for approaching API access and to not use as much base R but use custom packages with custom JSON handling. How to make Amazon Alexa skills

What's next for CEEDR

Expanding to other campuses, businesses and homes Putting energy consumption in more accessible terms for all users Target users to maximize savings through reduction in energy consumption Monitor population to determine whether some usage is unnecessary - survey and census

Share this project: