City as Systems: A Visualization of Economic "In's and Out's"

Visualizing the Hudson's economic "in's and out's"

Living in NYC and along the Hudson River, I see a lot of cargo ships coming in and out of the ports. The main questions at play when observing said ships are:

• How does the Hudson River act in this way, in our modern era of automobile vehicular transportation, airplane/aero vehicular transport, and train railways?
• When we see the cargo boats going through the Hudson and docking at the different ports, how much of that cargo belongs to what we've purchased/will purchase on a day-to-day basis?

Gathering data about the PANYNJ came from numerous different sources: the Bureau of Transportation Statistics, the PANYNJ Trade Statistics, U.S. Department of Transportation, etc. To composite an accurate and in-depth collection of data, I used AI (GPT 4o) to help sort through all different forms of data.

Research process

There were 5 main categories I sorted the cargo imported/exported into:

• Food & Agriculture
• Consumer & E-Commerce Goods
• Construction & Raw Materials
• Industrial/Machinery
• Other/Mixed Freight

The FAF uses 43 categories, and the PANYNJ Yearly Trade Reports do not detail every category/commodity that is listed/tracked by the FAF, but only the top 10 commodities. I had to extrapolate with the data I had gathered from the four main sources listed previously.

Utilizing AI for data gathering/data organization

If the information is displayed on a website (non-PDF), I collect the numerical data and reformat it into an excel sheet, which I then send to ChatGPT to reorganize and recategorize.

If the information is displayed via PDF, I inputted the source PDF itself and gave several strict criterion for GPT to look for and gather into a JSON file:

• Top 10 Containerized Imports (in TEUs)
• Top 10 Containerized Exports (in TEUs)
• Annual total imports (in TEUs)
• Annual total exports (in TEUs)
• Categorized into the 5 main categories determined previously.

Between the years of data collection (2018–2024), the units listed began to shift from TEU's (standardized containerized units) to metric tons. For this, I also required GPT to reconvert metric tons into TEU's for standardization. The general estimate formula for a TEU is: TEUs = (metric tons × 1000) ÷ 24