Will Fisher & Michael Chang

Hudson County

Will Fisher & Michael Chang

Introduction

Hudson County, named after the explorer Henry Hudson, was officially established in 1840. Just across the Hudson River from New York, its culture, economy and demographics are largely defined by the city it borders. In much the same way that Williamsburg has become an extension of the Lower East Side, the cities of Hudson County, in particular Hoboken and Jersey City, have become the outgrowths of downtown Manhattan. Hudson County is not just a commuter suburb, it is home to many of its attractions including Goldman Sachs's headquarters and the Liberty Science Center.

Hudson County is one of the densest counties in New Jersey. It is comprised of 12 counties, ranging from Bayonne on the lower harbor to Kearny, across the Passaic River from Newark, to Jersey City, a growing hub of financial services. In addition, in includes Union City, which is the most densely populated city in the United States at 53,000 people/sq. mile — although this is somewhat misleading since the Upper East Side is twice as dense but is not counted since it is not a city. We treat the county as a dense urban environment and design our PRT network in kind. The following table presents key statistics about Hudson County.[1]

Quick Facts / Hudson County / New Jersey
Population, 2008 estimate / 595,419 / 8,682,661
Language other than English spoken at home / 56.10% / 25.50%
Median household income, 2007 / $51,247 / $67,142
Private nonfarm establishments, 2007 / 13,308 / 243,350
Private nonfarm employment, 2007 / 220,144 / 3,661,679
Land area, 2000 (square miles) / 46 / 7,417
Persons per square mile, 2000 / 10,000 / 1,134

For more information about our county, please see the write-ups for Hudson County from years past. Next, we describe the current transportation network because our PRT network will have to interface with these current mass transit options.

Current Transportation Network

Hudson County is home to one of New Jersey's three light rail lines. The Hudson Berger Light Rail system (HBLR), which takes patrons along the main, north-south axis of Hudson County. HBLR starts in lower Bayonne and travels as far north as Port Imperial, roughly in line with 50th street in Manhattan.

As for heavy rail, there are three principal systems: the Northeast Corridor, Hoboken-terminated lines and PATH. On the Northeast Corridor (NEC) run trains from NJ Transit and Amtrak. NJ Transit also operates six different train services[2] out of its Hoboken terminal. Finally, PATH service connects Hoboken and other locations in Hudson County with both Newark and New York City. Hudson County is also home to important rail infrastructure including the Kearny Connection[3] and Secaucus Junction, both of which allow for greater ease of transportation for NJ Transit rail customers headed to Midtown Manhattan. See the picture below for a diagram of current service offerings.

More specifically, Hoboken Terminal, Journal Square, Exchange Place and Secaucus Junction are particularly important points of interest of public transit. In addition, NY Waterway offers ferry service from Jersey City, Hoboken and Weehawken to points in Manhattan.

Data

We should note that our business data is somewhat fuzzy. We realized early on that our business data was complete in the sense that the number of businesses listed is correct but the number of employees per business is either fabricated or just wrong. Since we couldn’t find a reliable database of businesses and the number of employees, we decided to take the existing set of businesses and scale the employees properly. We found the number of people employed in Hudson County (268,400),[4] divided by the number of businesses (4,866) and randomized each row’s employee count around that number (55) with a range of [2, 108] with an equal distribution. For the patrons, we took the following piecewise function where e is the number of employees of the business and p is the calculated number of patrons:

If (1 <= e <= 10): p = 80e, else if (10 < e <= 25): p = 60e, else: p = e.

This models a scenario where businesses of less than 10 employees are essentially small retail shops with high volume, medium size businesses are a larger form of retail, and any business with at least 25 employees has no “patrons” beyond people that visit the company for some reason (e.g. consultants). We think this is a good, generic model for business activity and the numbers we find average out to about 2 business-related trips per person per day.

For the school data, we extrapolated from the employee counts to get the student enrollment numbers. We found the average student-to-teacher ratio for Hudson County (14.7) and used this ratio as a scaling factor to find the number of attending students.[5]

We also included transportation points in our data file, including stations for heavy rail (NJ Transit, PATH) and ferry boarding points. We got the GPS coordinates from Google Maps and the ridership data from the respective websites.[6] Finally, we augmented the recreation data from a mere 5 sites to almost 50 using ReferenceUSA and found attendance data for the following types of recreational sites: amusement parks,[7] museums,[8] and malls.[9]

Finally, after the presentation, we went back to the data and made sure to delete all the businesses that were located outside our county. Some of the data points were falling on locations outside our county and were contributing to low trips-served percentages. After deleting these points, we observed an increase of ~3% in trips served.

Initial Network

Our initial network was a small network loop covering the area above I-495 (Lincoln Tunnel) up to the county's northern boundary. As a test, we found this useful because it allowed us to experiment with different track configurations in order to see which layout would allow the best amount of mobility with the least amount of track. Ultimately, we settled on a vertical loop configuration, similar to our initial network. See the figure below for a view of our initial network.

Final Network

What we have done for Hudson is create a vertical grid network. Essentially, the arterials of the PRT network are north-south tracks with switching interchanges along the way such that many small loops are formed (left figure). This design is based on Manhattan's street grid, which is largely successful because the blocks are small and so the necessary maneuvers (if the one-way street goes the wrong way, say) at the endpoints are minimal. We tried to cover Hudson County this way by placing stations approximately ¼ mile apart such that the spaces in between the north-south avenues are covered. In a handful cases, there were large swaths that were left without close access to a station and so we simply put a station in and connected it to the grid via interchanges. In the figure on the right, you can see that were the station named "center" not placed as such, there would be a large gap of unserviced area.

Our network also covers key points of interest including the financial district of Jersey City (Goldman Sachs building, Figure 1), Hoboken NJ Transit/PATH Station (Figure 2), I-495 Park & Ride (Figure 3) and the Newport Centre Mall (Figure 4). We feel it is important to design the network to accommodate popular venues in particular.

Figure 1: Goldman Sachs building.

Figure 2: Hoboken train station.

Figure 3: I-495 Park & Ride.

Figure 4: Newport Centre Mall.

We decided against modeling our system to accommodate the Hudson Bergen Light Rail system. Despite the fact that our specification says to make our PRT system complimentary to the light rail, we did the opposite. It is hard to justify designing a good PRT network around a service that is almost exactly the same. Light rail offers an automated for of mass transit serving a handful of popular points along one route. On the other hand, our PRT network will be faster (since it is not mass transit) and will have many more destinations, including all of the locations served by the HBLR. It will be ground level and automated. Not only would it be extremely difficult to model our network to be complimentary to light rail services, what would be the point? It is hard to imagine how light rail and PRT can be equal parts of a multimodal trip.

In addition to providing good service around residential areas, we also made a concerted effort to include stations near industrial areas with little residential zoning. This includes storage and shipping facilities, warehouses, etc. Hudson County is home to a disproportionately large industrial/services sector. This sector is centered in industrial zoning areas and we make sure to include these facilities so that the employees can get to and from work.

The following characterizes the design features of our network:

·  Our network links up with heavy rail stations, including Secaucus Junction and major PATH stations, ferry stations[10] and the sole NJ Transit bus Park & Ride location that we are aware of in Hudson County.[11]

·  Our PRT network lies on top of the existing street network and follows it closely with the liberal use of shape points. By designing our network as such, there would be no trespassing involved in constructing the guideways (i.e. no need to use eminent domain). In other words, it would be easier to pitch the idea since the construction would be simpler and less costly. In addition, we have placed some track along railways since they are straight and often the property of the public sector.

·  Our design does not use interchanges where only shape points are needed. If there is only one track coming in and one going out, this calls for a shape point, not an interchange. We obey this principle.

·  We have almost as many interchanges as we have stations. In Hudson County, this ratio is about approximately .9 interchanges per station. Since interchanges are free in our economic model, we use them liberally.

·  The network almost exclusively uses one-way guideways. You will find very few network edges that have track in both directions. This is important because one, it simplifies design and construction and two, it lowers costs as long as the network is still connected.[12] In a handful of places, one station is attached to network with a bidirectional guideway. We feel this is justified because the station platform can be positioned a la the drawing below. In this drawing, the blue lines represent the guideway, the green is the network and the red is the station platform (note the unique placement). We use this technique in place where it would make little sense to wrap around the station with tracks on either side. For an example in our network, see the figure on the right.

System Analysis

In looking at the station trip data, we noticed that a good number of stations serve zero trips. At first, this seemed odd to us, but on closer inspection it is clear why this is the case. In the below image, we see an example of a station where, ostensibly, we should be covering the locale's needs but because the census tracts are outside the station's quarter-mile circle, we end up with a trip-ends value of 0. In other instances, there are no official trip ends (since we didn't include the data) but service is still desired.

System Evolution

The PRT system is most likely to start growth in the most dense, urban areas. For Hudson County, this means Jersey City and Hoboken. Given the vast number of people that flow through these two municipalities and their densities, the cost per trip will be lower than anywhere else in the county. Ideally, however, the system would be fully funded and all the designed stations would be constructed. This has the advantage of offering a full countywide transportation system because what use is a mode of transport if it doesn't cover the desired destination?

Benefits

Among the obvious benefits are reduced traffic fatalities and reduced loss of time as a result of greatly diminished traffic congestion. Hudson County had 22 pedestrian deaths between 2006 and 2008.[13] We estimate that close the 7 lives will be spared if our PRT system is put into place. We can put economic values on the time savings and decreased fatality rate, but since they are not included in the spreadsheet model, we will leave this as is. If nothing else, these numbers would favor the construction of a PRT system since there are costs that will be saved by the residents of Hudson County.

Neighborhood Changes

We believe that the implementation of a successful PRT will result in higher property values in the affected areas and consequently, better recreational areas, better air quality, greater walkability and less crime. We claim that property values will go down because it has been shown that proximity to public transit increases home prices.[14] Assuming that cars are successfully replaced by our PRT system, it goes without saying that air quality will improve. Indeed, the overall environmental profile of our PRT system should be much better than that of an automobile-based system as long as the (power) grid is supplied by non-carbon sources or carbon-light fuels such as natural gas. As for the claims of reduced crime, we think that Jane Jacobs's notion of "eyes on the street"[15] will suffice to bring about reductions in crime. Jane Jacobs postulated that urban areas with more people walking around and fewer people in cars leads to safer areas. While traveling in a PRT pod is not conducive to this effect, walking to and from the station is key. Whereas cars can be parked at the home and used from end-to-end, PRT stations are located within walking distance of destinations. Thus, the extra walking required of our PRT system should be enough to produce the desired increase in walkability of the affected neighborhoods and in turn reduce crime in these areas.