Project 1
Student id: c12
To select another student or project, return to the Project Page.
<?xml version="1.0" encoding="UTF-8"?>
<?xml-model href="../schema_3302.rng" type="application/xml" schematypens="http://relaxng.org/ns/structure/1.0"?>
<!--the second line in the document associates the schema, so be sure not to change it-->
<DOC>
<docHead>
<!--required header includes metadata about the assignment (title, author, version)-->
<title>Writing Project 1 - Annotated Bibilography</title>
<author xml:id="c12">Brent Christy</author>
<version n="2" date="2016-07-13"/>
<!--note that the date must be YYYY-MM-DD for the document to be valid-->
</docHead>
<annotated_bib>
<problem_stmt> Every day throughout the city of Boston, thousands of commuters spend large
portions of their drive waiting in traffic. All it takes for traffic to build up is one
accident on a main street and cars will be backed up for miles. Part of this issue is
due to the volume of commuters entering and leaving the city during rush hour, but part
of the blame can also be placed on traffic control. Traffic control, specifically the
use of traffic signals, is fairly far behind today's technology<note type="instructor"
>specifically in Boston or in the U.S. or just in general?</note>. Although traffic
signals are monitored and timings are adjusted every five years, the results are still
not optimal. There needs to be a way for signals to utilize current technologies<note
type="instructor">automation? other specific tech?</note> in order to optimize
commute times and allow for people's already sparse free time to be spent how they like.
Along with people's free time being wasted, there is also a large amount of unnecessary
fuel consumption happening while waiting in traffic.<note type="instructor">good
point</note> Something needs to be done about traffic issues throughout high
population cities such as Boston. </problem_stmt>
<citation style="MLA" n="1"><title>Activity Report for FY 2013.</title> Rep. Boston Traffic
Management Center, n.d. Web.</citation>
<annotation>
<background type="source"> The Boston Traffic Management Center produces an annual
activity report in order to publish the statistics and usage of its department. </background>
<summary type="general">Currently, the Traffic Management Center (TMC) <q>"has computer
control of 544 out of the 845 traffic signals operated by the Boston
Transportation Department (64%)"(1).</q><note type="instructor">we didn't talk
about this in class, but for purposes of display, the q element adds quotation
marks automatically</note> The department made 31,000 real-time manual changes
to the timing of traffic signals based on current traffic conditions. The department
also owns 195 traffic monitoring cameras throughout the city.</summary>
<relevance type="application">These statistics show that the Boston Transportation
Department has part of the infrastructure to implement an automated real-time
traffic control system. The automated system would be able to make the 31,000 manual
changes that employees made throughout the year by itself and reduce
traffic.</relevance>
</annotation>
<citation style="MLA" n="2"><author n="1">DeMarco, Peter</author>. <title>"Exploring the
Science behind Traffic Lights."</title> Boston.com. N.p., 09 May 2010. Web. 13 July
2016.</citation>
<annotation>
<background type="source">An article published by Boston.com discusses the current
technology used in traffic signals in the Quincy area.</background>
<summary type="general">Jack Gillon, a traffic engineer in Quincy, explained how he has
total control of the lights from a central location. He also quickly showed how
sensors detect cars at traffic signals.</summary>
<summary type="approach">Most traffic signals have an electrically charged wire
underneath the pavement. When a car passes over this wire, a change in magnetic
field is measured and sent to the control box of the light. This tells the light
that there is a car waiting to go through the intersection or make a turn.</summary>
<relevance type="application">This technology could be implemented in an automated
traffic light system by sending the readings from the charged wire to a central
network. The central network could then take in these readings and account for the
incoming traffic at the next lights that the car would encounter.<note
type="instructor">so the system would develop (through inferential statistics or
something) an understanding of the typical routes of cars and the current
traffic to automate?</note></relevance>
</annotation>
<citation style="MLA" n="3"><author n="1">Gilmore, John F.</author>, <author n="2">Khalid J.
Elibiary</author>, and <author n="3">Naohiko Abe</author>. <title>"Neural Network
System for Traffic Flow Management."</title> AAAI Technical Report (1993): 85-95.
Web.</citation>
<annotation>
<background type="author"> John Gilmore along with Khalid Elibiary of the Georgia Tech
Research Institute produced a paper for Association for the Advancement of
Artificial Intelligence regarding the use of neural networks to control traffic in
1993 with the assistance of Naohiko Abe of Mitsubishi Heavy Industries in Japan. </background>
<summary type="general"> Gilmore and his associates go into depth about how neural
networks can be implemented into a traffic control system to improve traffic
conditions. They introduce the idea of using the Hopfield neural network model in
order to allow the network to learn and build off of its knowledge of previous
congestion situations to improve traffic conditions.<note type="instructor"
>fascinating</note>
</summary>
<summary type="approach">The group forms an energy function which allows them to assign
lights and traffic conditions different weights. Higher weighted traits, such as how
busy an intersection generally is, allow lights to have a higher priority and change
faster. </summary>
<relevance type="application"> The usage of a neural network as described by Gilmore,
Elibiary, and Abe can be used as a base for a city wide traffic control system,
where the system learns and adapts based on previous traffic data. This could cut
out a large amount of traffic by optimizing the traffic flow.<note type="instructor"
>really solid methodology source here</note>
</relevance>
</annotation>
<citation style="MLA" n="4"><author n="1">Santi, Paolo</author>, and <author n="2">Carlo
Ratti</author>. <title>"Revisiting Street Intersections Using Slot-Based
Systems."</title> PLOS ONE. N.p., n.d. Web. 13 July 2016.</citation>
<annotation>
<background type="author">Paolo Santi of the Senseable City Lab and Carlo Ratti of MIT
outlined a plan to eliminate the need for traffic lights at all using automation in
cars.</background>
<summary type="general">For Santi's system, called a slot-based system, to work, cars
need to be autonomously operated. The cars need to be able to communicate with each
other to determine where they are on the road and what their planned route is at
each light. The cars would then decide amongst themselves which car is turning when
with no need for traffic lights. </summary>
<relevance type="application">If all cars could communicate and turn without interfering
with each other, it would eliminate the need for stop lights and would greatly
decrease traffic given that the cars will harmoniously move through the
streets.</relevance>
<relevance type="value_stmt">Although this seems like the optimal solution, it would
first require that all cars be autonomous, so this would not be a solution for the
near-term.<note type="instructor">good point--but if we think this is an
eventuality then would that affect whether or not the infrastructural investment
in a neural traffic network would be worth it? Or how likely stakeholders are to
support your approach?</note></relevance>
</annotation>
<citation style="MLA" n="5"><author n="1">Srinivasan, Dipti</author>, <author n="2">Min Chee
Choy</author>, and <author n="3">Ruey Long Cheu</author>. <title>"Neural Networks
for Real-Time Traffic Signal Control."</title> IEEE Trans. Intell. Transport. Syst.
IEEE Transactions on Intelligent Transportation Systems 7.3 (2006): 261-72.
Web.</citation>
<annotation>
<background type="author">Dipti Srinivasan and her colleagues explore the use of
real-time neural networks in order to assist with traffic control in this paper for
an Institute of Electrical and Electronics Engineers journal on Intelligent
Transportation Systems. </background>
<summary type="general">Srinivasan uses the concept of a neural network to take in
real-time traffic statistics and adjust traffic lights accordingly. The current
conditions of traffic are available to the system and can be included in
calculations made to determine how long to have lights turned green.</summary>
<relevance type="value_stmt">Her results showed that the implementation of a real-time
neural network can allow a greater number of cars to pass through congested
intersections.<note type="instructor">could argue for summary @type
'interpretation' here, but this was a sticking point, I think, for all of us in
developing the schema</note>
</relevance>
<relevance type="application">Dipti's real-time neural network could be applied in a
city like Boston where sensors could be installed to acquire real-time data used in
the program to control the lights. Traffic control based on current conditions would
produce much better traffic flow than average timed lights.<note type="instructor"
>is their approach similar to your second source? how do these sources connect
(or how do you see them speaking to each other)?</note></relevance>
</annotation>
</annotated_bib>
<docReview>
<reviewer type="instructor" xml:id="kgs"/>
<report type="instructor" resp="#kgs">
<p>Excellent job here. You have a good mix of topical and methodological sources that
apply to your problem. One thing you may want to think about is what an overhaul of
the traffic system might look like logistically--who are the stakeholders? who do
you need support from, etc.? Thinking in terms of the Leavy reading from last week,
it would be interesting to think about how a multi/inter/transdisciplinary team
might approach this problem from the technical, infrastructural, and political/labor
sides.</p>
</report>
</docReview>
</DOC>