Zac Rego's GIS

Assignment 4

Agate Bike Counter Story

1. The environmental data seems to have no impact on the bicycle counts. There is a very good correlation between population (enrollment) and bicycle count, however. There is also a very interesting trend within the counter data – a saw tooth pattern in the hourly counts. I hypothesize that this can be explained by a more intimate look at enrollment data, and the fact that only the southern half of the counters are working. Firstly, the fact that the southern half of sensors is the only half that works means that much more incoming traffic will be counted than outgoing traffic, as incoming riders are more likely to be on the right side of the street, which is the southern half. Secondly, after a brief look at the class registration menu for the coming fall, it seems that classes longer than 50 minutes start more frequently at certain hours than others. 10 am, 12 pm, and 2 pm are more common for classes longer than 50 minutes than 11 am, 1 pm, and 3 pm. When looking at the saw tooth pattern, the time stamps for 10 am, 12 pm, and 2 pm are the valleys, and 11 am, 1 pm, and 3 pm are the peaks. This pattern is in agreement with my hypothesis, because from 10-11 am, the majority of students arriving on campus will be for a 11:00-11:50 class, whereas from 11-12 pm, students in 12:00-12:50 and 12:00-1:20 will be arriving. This is reflected in the peaks and valleys of the graph.

2. The sensors measuring bikes are electromagnetic inductance sensors. These measure metals passing over them by detecting a change in the EM field above the sensor. Bikes have a very small amount of metal in them and are not very wide, so many sensors are needed to accurately count the bikes that pass overhead. This challenge is further made difficult by the layer of bricks between the sensors and the road surface.

3. The data is getting from the sensor network to us over a wifi network. The sensors send their data to a computer which analyzes and formats it, then sends it over wifi to a server where it can be accessed.

4. I chose to use mostly dual axis line graphs because the data is continuous and I wanted to compare bicycle counts to one variable at a time. I chose the order of my story in order to show firstly what was not affecting counts and then what was, ending with “note the sawtooth pattern” as a lead in to these questions.

Eugene Bike Counter Dashboard

1. The Eugene bike counter data seems to trend in the exact opposite way that the UO bike counter data does. Rather than correlate with campus population, the numbers seem to better correlate with the weather. There is a rough correlation between average daily temperatures and bicycle counts, and another rough correlation between season and bicycle counts. There is less variance in weekday counts during the summer months, which seems to indicate that biking is less work related. There is also a large drop in counts during the winter months, while the highest counts occur during fall and spring.

2. The sensors that gather bicycle counts for the city are different than the one used for the intersection on 13th and Kincaid. While 13th and Kincaid uses an induction loop, and counts continuously, remotely sending the data, the city used pneumatic tubes and sampled intersections for only a certain number of days in a row. Every time a bicycle rode over a tube, a pressure wave was sent to the sensor where it was registered as a bicycle. These sensors are able to filter out other kinds of traffic based on signatures in the pressure waves. The data also had to be gathered on site, rather than being able to access it through the internet.

3. This data had to be gathered on site and manually uploaded into a computer. Once all the data was gathered for each different intersection, it was manually compiled so that it could be manipulated by programs such as excel or tableau.

4. I chose to use dual axis line graphs because they are simple, straightforward visualizations that allow me to compare one variable to another. For the seasonal graph, I wanted each season to be discrete so that I could also look at weekday variance by season.

Final Project Proposal

Parking in dense urban areas:
Parking downtown can be a difficult, stressful experience. Relying on old fashioned meters and payment kiosks makes the situation even more complicated, especially when one does not have any change, or puts the paper slip on the wrong side of the window. I am suggesting a new infrastructure which makes both parking and enforcement of parking time limits/zoning easier and more reliable. This infrastructure would be similar to that on an automated toll bridge, where a transponder is placed in a car and read by a stationary reader. Using this system automates payments, so a person never has to carry change or make any payments at the time of parking. It would also improve the efficiency of parking enforcement, as an enforcer would be notified when a vehicle’s time has expired, or when a vehicle without the proper zone permissions parks in an unauthorized zone. This technology could also be placed in the entrances or exits of parking garages, as a faster method to complete transactions.

1. Context – Smart City
a. My project fits into the context of a smart city because it integrates an area of infrastructure into a network, which automates much of the process. This project will use automatic sensors to gather data from unique transponder signals placed in cars, and make automated decisions. Not only does this project make the process of parking and ticketing easier, but it could be used to gather spatial data over time, which could be used in a policy making process.
2. Sector – Transportation/Mobility
a. The primary purpose of this project is to remove inefficiencies in the current parking infrastructure. This immediately impacts transport and mobility because it will: a. affect how commuters choose modes of transportation, b. increase the efficiency of the ticketing system, and c. allow a city to better regulate parking spaces, by utilizing special zones or time limits. It will also put a large number of vehicles on the same network, which will allow the city to make future decisions regarding parking.
3. Purpose – Goals and Objectives
a. The goal of this project is to making parking in dense, popular areas (such as downtown areas) easier, and at the same time make the ticketing process easier. It will remove the need to perform any transactions at the time of parking, and patrolling of police parking enforcers will no longer be necessary, as they will be notified of exactly which spaces have violators in them.
4. Scope/Scale
a. The realistic scale of this project is a dense urban area. It would be unnecessary to extend this project beyond such an area because both parking and enforcement are made easier by the smaller density of cars. It would also be much too costly to spread the project across a large city, especially one with large residential areas. Further, in these areas, the infrastructure required to operate this system would be an unsightly installation.
5. Sensor Ontology
a. http://www.atlasrfidstore.com/skyetek-supernova-rfid-reader/
b. http://www.atlasrfidstore.com/rearview-mirror-rfid-hang-tag/
c. These two pieces of technology are the RFID readers and RFID tags that will be used when parking. The RFID tags can reach a maximum of about 18 ft, which should be more than enough for a city block application, and are passively powered, which means they receive all their power from the signal coming from the RFID reader. The reader only requires a 5V USB power source. These are also consumer products; a city should be able to buy these in great quantities without the branding, packaging, and storefront interface.
6. Sampling
a. This project will require 1 RFID reader per parking space. I originally wanted to use 1 reader for multiple spaces, but the reader would not know where the car was parked, and this is a crucial aspect of the project. The readers will be installed in fixed positions, and be programmed to only read within a very small range, so that passing cars are not picked up. They would also be programmed to send out a signal every 5-10 minutes, so that they can recognize when a certain car has left the space. Depending on the needs of the city, the tags can be programmed so that the unique ID is linked just to a payment account, or to other personal information as well, such as name and address. This would allow the city to gather much more data about its citizens’ transportation habits, but would present a big privacy risk.
7. Network
a. The RFID readers will need to be connected to a central server or network of servers where the data is processed. Because of the large number of sensors, each street would need its own node, or multiple nodes, to process the incoming data. This “preprocessed” data would then be more easily sent to the central processing stations. This could be done wirelessly or with an actual infrastructure of wires. Either way, there would need to be a live feed, so that parking enforcers would have access to the data, and could act on violators.
8. Data
a. This system will generate multiple kinds of data. First, the readers will read the RFID tag, which will have a unique ID in them. This unique ID, along with the location of the reader, will be sent to a node or a server. This will start multiple processes, the first is that the ID of the space will be read, so that an appropriate timer can start, and to verify that the unique ID has permission to park in that space. If the car stays in the space for too long, or does not have permission to park, parking enforcement will be notified so that a ticket can be issued. If the car does have permission and stays in the space for an acceptable amount of time, when it has left, the account linked to the ID will be charged the appropriate amount. The rest of the information associated with this ID may also be read and stored. The central servers will keep track of how many times each reader is used, so that frequency data for parking spaces can be gathered and later analyzed. Only the city will have access to the data, and data will not have to be stored for a very long time before it can be consolidated and stored in smaller forms.
9. Budgets and Money
a. The consumer prices of this technology are high. Each RFID tag costs $24, and each reader costs $499, and this does not include the price of setting up, operating, maintaining, and staffing the network. I assume that the wholesale price of this equipment will be less, but ideally traffic enforcers will more efficiently be able to ticket parking offenders, which should increase profits and help pay for the new system.
10. Visualization
a. Because each RFID reader will have its own unique ID, and the locations of each of these IDs will be known, this data can be easily spatially visualized. The city can track how often a space is used or violated, allowing adjustments to easily be made. If RFID tags include address information for drivers, the city will also be also be able to see where each car is coming from, which could be useful in other planning applications. The data will also need to be spatially visualized for the use of parking enforcers. This could be in the form of a simple address punched into a mapping application.
11. Concerns – Privacy, Security, and Morality
a. There are a few privacy and security concerns with this project. The first being that unique IDs will be linked to payment or bank accounts, which if accessed by a malicious party could be jeopardized. This could be remedied by allowing users to maintain a payment account which they supply through bank or credit payments. Secondly, users may not want personal information linked to these IDs. Drivers may not want the city to know their driving and parking habits. This city would most likely have to advertise the anonymity of the users to gain their trust. Another concern is that it would take a great deal of effort to get every driver integrated into this system, and there would never be 100% of drivers using this system, especially with the introduction of travelers. In this case, the current payment methods would have to remain installed to allow these drivers to pay for their spaces.