-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
310 lines (306 loc) · 18.3 KB
/
index.html
File metadata and controls
310 lines (306 loc) · 18.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta content="width=device-width, user-scalable=no, initial-scale=1.0, maximum-scale=1.0, minimum-scale=1.0"
name="viewport">
<meta content="ie=edge" http-equiv="X-UA-Compatible">
<title>Interagency Work Zone Traffic Data Modeling and Analysis </title>
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="src/js/bootstrap.min.js"></script>
<script src="src/js/bootstrap.bundle.min.js"></script>
<link href="src/css/bootstrap-grid.min.css" rel="stylesheet">
<link href="src/css/bootstrap.min.css" rel="stylesheet">
<link href="src/css/main.css" rel="stylesheet">
</head>
<body data-offset="150" data-spy="scroll" data-target="#sidebar">
<div class="container-fluid">
<div class="row" id="body__row">
<div id="nav_fix-top">
<img alt="navigation-icon" id="nav__open-btn" onclick="nav_show()" src="src/image/nav.png">
<button id="nav__close-btn" onclick="nav_close()">Close</button>
</div>
<nav class="col-md-3 col-lg-3" id="sidebar">
<ul class="nav__content-wrapper nav__main">
<li class="nav-item">
<a class="nav-link .active" href="#main__cover-page" id="nav__home-btn">Home</a>
</li>
<li class="nav-item">
<a class="nav-link" href="#main__overview">Overview</a>
</li>
<li class="nav-item">
<a class="nav-link" href="#main__data">Project Data</a>
</li>
<li class="nav-item">
<a class="nav-link" href="#main__methodology">Methodology</a>
<ul>
<li><a class="nav-link nav-subsection" href="#methodology__wrangling">Data Wrangling</a></li>
<li><a class="nav-link nav-subsection" href="#methodology__clustering">Clustering</a></li>
</ul>
</li>
<li class="nav-item">
<a class="nav-link" href="#main__results">Results</a>
<ul>
<li><a class="nav-link nav-subsection" href="#results__dashboard">Dashboard</a></li>
<li><a class="nav-link nav-subsection" href="https://workzone-collision-predict.herokuapp.com/"
target="_blank">Web
application</a></li>
</ul>
</li>
<li class="nav-item">
<a class="nav-link" href="#main__about">Team</a>
</li>
<li class="nav-item">
<a class="nav-link" href="https://github.com/workzone-collision-analysis/capstone" target="_blank">Github</a>
</li>
<li class="nav-item">
<a class="nav-link" id="report_download" download="report.pdf" href="src/report.pdf">Report</a>
</li>
</ul>
</nav>
<main class="col-md-12 col-lg-9 ml-sm-auto" id="main">
<section class="container-fluid main__section" id="main__cover-page">
<div id="logos">
<img alt="NYU CUSP LOGO" src="src/image/cusp.png">
<img alt="HDR" src="src/image/hdr.png">
</div>
<h1>Interagency Work Zone<br> Traffic Data Modeling and Analysis </h1>
<h4>NYU | Center for Urban Science and Progress</h4>
<h4>Capstone Project | 2020 </h4>
</section>
<section class="container-fluid main__section" id="main__overview">
<h2>Overview</h2>
<h3>Background</h3>
<p>Road construction events are a necessary part of keeping road infrastructure in good condition but
can pose significant safety problems when implemented. Temporary work zones for roadway
constructions have the potential to significantly impact mobility and safety for all roadway users.
An increase in the number of people using local streets because of work zone diversion plans may
increase the likelihood of crashes, including crashes involving vulnerable populations (e.g.,
cyclists, seniors, and individuals with disabilities). To aid transportation authorities in the city
of New York better understand the extent of mobility impacts associated with work zones, we propose
a clustering approach to predict the probability of a vehicle collision occurring in the proximity
of a road construction event (i.e. work zone). </p>
<h3 id="problemState">Problem Statement</h3>
<div>
<h1>Can we use characteristics of streets and roadway construction events to predict collision
probability in future work zones? </h1>
</div>
<h3>Scope</h3>
<div class="row">
<div class="col-8 col-sm-5 col-md-4 col-lg-4 scope__column">
<img alt="Dataset Diagram" src="src/image/database.png">
<h4>Data Wrangling</h4>
</div>
<div class="col-8 col-sm-5 col-md-4 col-lg-4 scope__column">
<img alt="Clustering Diagram" src="src/image/clustering.png">
<h4>Clustering</h4>
</div>
<div class="col-8 col-sm-5 col-md-4 col-lg-4 scope__column">
<img alt="Dashboard Diagram" src="src/image/dashboard.png">
<h4>Dashboard</h4>
</div>
<div class="col-8 col-sm-5 col-md-4 col-lg-4 scope__column">
<img alt="Dashboard Diagram" src="src/image/webapp.png">
<h4>Predictive Web Application</h4>
</div>
</div>
<p>This project proposes using a k-means clustering approach to predict the probability of a vehicle
collision occurring in the proximity of a work zone. The proposed clustering method is applied to
over 20,000 construction and emergency construction events of relatively short duration in New York
City to identify types of work zones that may present greater safety risks. The results of this
project enables practitioners to employ appropriate mitigation strategies during project programming
and develop effective transportation management plans. </p>
</section>
<section class="container-fluid main__section" id="main__data">
<h2>Project Data</h2>
<p>This project is unique in that all data sources used in this research are publicly available
datasets. The team did not have access to a single, unified data source that provided information
about the location and timing of construction work zones within NYC. Instead, the team brought
together disparate datasets to produce an approximation of work zone events and their attributes
(road type, duration, length, etc.). The information below shows the sources and features taken from
each dataset to produce a single dataset of construction events. </p>
<div class="row">
<div class="col-8 col-sm-5 col-md-4 col-lg-4 data__column">
<div class="data__column-icon">
<img alt="road construction" id="icon_511" src="src/image/511.png">
</div>
<h4>511 Events</h4>
<span>Geometry: Point</span>
<ul class="data__column-attribute">
<li>Location,</li>
<li>Duration,</li>
<li>Season</li>
</ul>
</div>
<div class="col-8 col-sm-5 col-md-4 col-lg-4 data__column">
<div class="data__column-icon">
<img alt="road construction" src="src/image/construction.png">
</div>
<h4>Street Closure</h4>
<span>Geometry: Line</span>
<ul class="data__column-attribute" id="sc__attr">
<li>Length</li>
</ul>
</div>
<div class="col-8 col-sm-5 col-md-4 col-lg-4 data__column">
<div class="data__column-icon">
<img alt="crash symbol" id='crash_icon' src="src/image/crash_icon.png">
</div>
<h4>Crash</h4>
<span>Geometry: Point</span>
<ul class="data__column-attribute" id="crash__attr">
<li>Number of crashes</li>
</ul>
</div>
<div class="col-8 col-sm-5 col-md-4 col-lg-4 data__column">
<div class="data__column-icon">
<img alt="crash symbol" src="src/image/lion.png">
</div>
<h4>LION</h4>
<span>Geometry: Line</span>
<ul class="data__column-attribute">
<li>Roadway type,</li>
<li>Street width,</li>
<li>Posted speed</li>
</ul>
</div>
</div>
</section>
<section class="container-fluid main__section" id="main__methodology">
<h2 id="section3">Methodology</h2>
<div id="methodology__wrangling">
<h3>Data Wrangling</h3>
<p>The data sources were produced by various entities and used different geometric representations.
Therefore, there was a need for a standardized street network onto which data sets, namely LION,
crashes, and work zones, could be joined. SharedStreets – a tool that creates a shared reference
system for disparate street networks – provided a solution to connect the data. Each record in
the LION, collisions and WZ data was assigned a ‘SharedStreets geometry id’. Based on this id,
characteristics of streets, collisions, and work zones were linked. The data was filtered to
include only durations less than 24 hours, and number of crashes happening within 900 feet of a
work zone was calculated. Then the Street Closures data was joined to get the length attribute
in feet. </p>
<div class="methodology__image">
<img alt="flow chart of the project" src="src/image/flowchart.png">
</div>
</div>
<div id="methodology__clustering">
<h3>Clustering</h3>
<p>With the spatial datasets wrangled into a common reference system, the team employed a k-means
clustering algorithm. This approach was selected to provide a predictive framework to understand
the probability of a crash occurring in the proximity of a work zone given certain
characteristics. The clustering approach also provides the opportunity to identify different
“cohorts” of work zones and identify work zones that may have a higher crash probability. Two
separate clustering attempts were made – one with a larger dataset that did not contain a length
attribute and one smaller subset of records that were able to be matched to permit data that
contained a length attribute. The collision probability was then calculated by dividing the
number of WZs witnessing a crash by the total number of WZs for each cluster. </p>
<div class="methodology__image">
<img id='clustering_img' alt="clustering" src="src/image/cluster_method.png">
<img id='clustering_img_mobile' alt="clustering" src="src/image/cluster_method_mobile.png">
</div>
</div>
</section>
<section class="container-fluid main__section" id="main__results">
<h2 id="section4">Results</h2>
<h3>Clustering</h3>
<p>The silhouette analysis was used to choose an optimal value for the number of clusters k=4. The
predicted probability (using the train set) of a crash happening in each of the four clusters is
16%, 13%, 14%, and 16% respectively.
<a href="https://workzone-collision-analysis.github.io/capstone/dashboard/" target="_blank">
An interactive dashboard</a> was designed to serve as a tool for
transportation authorities to explore the public safety risks associated with historical work events
as well as the historical crash rates on all New York City roads and intersections.
<a href="https://workzone-collision-predict.herokuapp.com/" target="_blank">A predictive web
application</a> was also built off the results of the clustering methodology to inform the
planning of
multiple new construction events simultaneously. It would aid decision making to avoid situations
with high crash risk.</p>
<div id="results__dashboard">
<h3>Dashboard</h3>
<iframe src="https://workzone-collision-analysis.github.io/capstone/dashboard/"></iframe>
</div>
</section>
<section class="container-fluid main__section" id="main__about">
<h2>Team</h2>
<p>This project was completed as part of the 2020 Capstone process for the Center for Urban Science and
Progress (CUSP) at NYU. The team would like to thank capstone sponsor HDR and their partners for
their help in the execution of this research. </p>
<ul class="team__student">
<li class="student__column">
<div class="student__circle">
<span>Collier, John</span>
<div class="team__link">
<a href="https://www.linkedin.com/in/john-collier-6557115b/">
<img alt="linkedin link" src="src/image/linkedin.png">
</a>
<a href="https://github.com/johncollier">
<img alt="github link" src="src/image/github.png">
</a>
</div>
</div>
</li>
<li class="student__column">
<div class="student__circle">
<span>Han, Seunggyun</span>
<div class="team__link">
<a href="https://www.linkedin.com/in/seunggyunhancodes/">
<img alt="linkedin link" src="src/image/linkedin.png">
</a>
<a href="https://github.com/Aete">
<img alt="github link" src="src/image/github.png">
</a>
</div>
</div>
</li>
<li class="student__column">
<div class="student__circle">
<span>Jaber, Linda</span>
<div class="team__link">
<a href="https://www.linkedin.com/in/lindajaber/">
<img alt="linkedin link" src="src/image/linkedin.png">
</a>
<a href="https://github.com/lindaJaber">
<img alt="github link" src="src/image/github.png">
</a>
</div>
</div>
</li>
<li class="student__column">
<div class="student__circle">
<span>Singh, Akhil</span>
<div class="team__link">
<a href="https://www.linkedin.com/in/akhil-k-singh/">
<img alt="linkedin link" src="src/image/linkedin.png">
</a>
<a href="https://github.com/akhilksingh">
<img alt="github link" src="src/image/github.png">
</a>
</div>
</div>
</li>
</ul>
<h3>Mentors</h3>
<ul class="team__mentor">
<li class="mentor__column">
<div class="mentor__circle">
<span>Ozbay, Kaan</span>
<div class="team__link">
<a href="https://www.linkedin.com/in/kaan-ozbay-4805413/">
<img alt="linkedin link" class="linkedin" src="src/image/linkedin.png">
</a>
</div>
</div>
</li>
<li class="mentor__column">
<div class="mentor__circle">
<span>Khan, Junaid</span>
</div>
</li>
</ul>
</section>
</main>
</div>
</div>
<script src="src/js/nav.js"></script>
</body>
</html>