Inspiration
Energy drink prices, hours and payment options on campus aren't published in one place. Before we could build anything for students, we needed that scattered information in a structure we could explore and question. Databricks is where we did that.
What it does
Databricks is the data foundation of TAMU Energize, an interactive map of energy drink prices in College Station. We structured our first version of the data there as two datasets, campus_buildings and energy_drink_catalog, covering 7 buildings, 8 venues, 5 brands and 27 drink options.
Our dashboard, "Aggie Map: Campus Buildings & Energy Drinks," gave us one view of:
- Counters for buildings, locations, brands and drink options
- A Building & Venue Guide with building and venue hours side by side
- Average energy drink price by building
- Locations by venue type
- A full Energy Drink Catalog
The final app carries that dashboard forward, with its charts and tables calculated live from the same data students see on the map.
How we built it
We separated places from products: a building has venues, and a venue sells drinks. We then built the dashboard to see what the data could tell us before designing the app.
Next, we exported the dashboard and converted both datasets into the data file that powers our map, keeping the same structure. We rebuilt each widget in JavaScript so the app needs no live connection to a warehouse, and students can open it from a free GitHub Pages link.
Finally, we used the Databricks data as a checklist for field work. We traveled around campus, checked locations in person, and recorded 22 timestamped prices across 6 vendors in 5 buildings, along with whether each vendor takes Dining Dollars.
Challenges we ran into
- Turning a dashboard into an app. A dashboard export isn't application data, so we had to convert it into a format the map could use.
- Errors in our own data. The dashboard labelled ILCB as the Interdisciplinary Life Sciences Building, which is a different building. We corrected it using OpenStreetMap.
- Hours that didn't line up. Some venues had a single hours window that didn't fit their building's schedule. The Pavilion's Aggie Express was listed as open at times its building was closed.
- Knowing what to trust. A clean-looking dashboard can make unconfirmed numbers feel certain, so we marked which prices were later seen in person.
Accomplishments that we're proud of
- A simple two-table structure that carried from our first dashboard all the way into the live app.
- Rebuilding the dashboard inside the app, so its numbers can't drift from what students see on the map.
What we learned
- Exploring data in a dashboard first made the product better. Seeing building and venue hours side by side made it clear that the app needed open and closed status for each vendor.
- A good data model saves work later, because ours never needed restructuring.
What's next for Databricks - TAMU Energize
- Load our in-person observations into Databricks as a new table, so the dashboard reflects what we saw on site.
- Record repeat visits to build a price history and track changes over time.
- Add dashboard views comparing vending machines with stores, and Dining Dollars coverage across campus.
- Compare observed prices with listed prices to measure how far listed data drifts from reality.
Built With
- databrick
