Jersey City Population Growth: "Jersey City will be the largest city in New Jersey by the end of 2016" - Mayor Fulop

• Major renovations are coming down the pipe for the PATH train

• Major renovations are coming down the pipe for the PATH train
• Expansion to Newark Airport

• Major renovations are coming down the pipe for the PATH train
• Expansion to Newark Airport
• Port Authority plans to expand ridership capacity by 20% by 2017

Given turnstile data in 15 minute blocks, determine when and where build-ups occur.

Challenges:

• 15 minute blocks (several trains per block)
• No tracking of individuals (no swipe-in/swipe-out)
• Transfers

To get the passenger flux, I decided of some limitations:

• Nobody is going from 33rd St. to WTC (or the other way) by the PATH

To get the passenger flux, I decided of some limitations:

• Nobody is going from 33rd St. to WTC (or the other way) by the PATH
• Nobody is taking the PATH to go from 33rd St. to another stop on the yellow line in Manhattan apart for Christopher st.

To get the passenger flux, I decided of some limitations:

• Nobody is going from 33rd St. to WTC (or the other way) by the PATH
• Nobody is taking the PATH to go from 33rd St. to another stop on the yellow line in Manhattan apart for Christopher st.
• People that change from red line to yellow line do so at JS; not at Grove
• Kept only the weekday data

To get the passenger flux, I decided of some limitations:

• Nobody is going from 33rd St. to WTC (or the other way) by the PATH
• Nobody is taking the PATH to go from 33rd St. to another stop on the yellow line in Manhattan apart for Christopher st.
• People that change from red line to yellow line do so at JS; not at Grove
• Kept only the weekday data
• Averaged the number of entry and exit in each station accross the week.

For example one passenger arrived at a station at 9 am:

• Find all the trains stoping in this station between 9 and 9:30.

For example one passenger arrived at a station at 9 am:

• Find all the trains stoping in this station between 9 and 9:30.
• Find all the possible exits for all these trains.

For example one passenger arrived at Newport at 9 am:

• Find all the trains stoping at Newport between 9 am (time of the entry) to 9:30am (+ 30 min).
• Find all the possible exits for all these trains.
• Account for the possibilities of a transfer.

For each exit:

• Look at all the possible entry (from the previous step)

For each exit:

• Look at all the possible entry (from the previous step)
• Assign passenger to each possible entry stations in proportion of the number of passenger in these entry station

• After that, most passenger are assigned to a train.
• Some passenger have only entry or exit

• These passengers seems to be an anomaly in the data

• These passengers seems to be an anomaly in the data
• It would be important to explain this problem

• These passengers seems to be an anomaly in the data

• It would be important to explain this problem

• We added to train:
  • taking the time an station they were entering/leaving
  • in proportion to were assigned passengers are going/coming from

• Inputs:
  • train schedule
  • turnstile data
  • likely rider behavior

• Goals for visualization:
  • station level
  • system level

• Method to estimate ridership behavior from 15 min bins
• Wireframe visualization of station-specific and system-wide congestion