1
00:00:00,269 --> 00:00:01,102
(somber music)

2
00:00:01,102 --> 00:00:03,769
(logo whooshes)

3
00:00:07,410 --> 00:00:10,320
- [Presenter] A quarter of
the 38,000 fatal crashes

4
00:00:10,320 --> 00:00:13,770
nationwide occur on curved roadways.

5
00:00:13,770 --> 00:00:17,820
Curves also play a significant
role in Texas roadway safety,

6
00:00:17,820 --> 00:00:20,610
with around 22% of fatal crashes

7
00:00:20,610 --> 00:00:24,120
over a seven year period
being curve-related.

8
00:00:24,120 --> 00:00:26,250
Even at those numbers, it's possible

9
00:00:26,250 --> 00:00:28,170
that many of these types of crashes

10
00:00:28,170 --> 00:00:30,690
are categorized incorrectly when entered

11
00:00:30,690 --> 00:00:34,290
into the state's Crash
Records Information System.

12
00:00:34,290 --> 00:00:37,170
This means that the role
of curves in accidents

13
00:00:37,170 --> 00:00:39,060
may be underestimated.

14
00:00:39,060 --> 00:00:42,210
Researchers at the Center
for Transportation Research

15
00:00:42,210 --> 00:00:44,520
develop methods for better understanding

16
00:00:44,520 --> 00:00:48,210
curve characteristics and
their impact on crash risk

17
00:00:48,210 --> 00:00:50,310
and severity in Texas.

18
00:00:50,310 --> 00:00:52,200
- There are times that
we need to sample data

19
00:00:52,200 --> 00:00:53,997
to find ways to improve it,

20
00:00:53,997 --> 00:00:56,760
and that's what this
research was trying to do,

21
00:00:56,760 --> 00:00:59,850
is looking for a way to
check and then improve

22
00:00:59,850 --> 00:01:00,810
curve-related data

23
00:01:00,810 --> 00:01:03,133
inside the Crash Record
Information System.

24
00:01:03,133 --> 00:01:08,040
- The purpose of Project 0-7050

25
00:01:08,040 --> 00:01:13,040
is to develop our automated
methodological procedure

26
00:01:13,350 --> 00:01:17,460
that can effectively
improve the identification

27
00:01:17,460 --> 00:01:22,110
of curve-related crashes
in the CRIS database.

28
00:01:22,110 --> 00:01:27,090
Using curve-related
information, the CRIS database,

29
00:01:27,090 --> 00:01:31,890
and the Highway Curves GIS Layer,

30
00:01:31,890 --> 00:01:36,510
curve-related crash
misclassifications were categorized

31
00:01:36,510 --> 00:01:38,400
into six types.

32
00:01:38,400 --> 00:01:41,700
Based on the six types
of misclassifications,

33
00:01:41,700 --> 00:01:46,700
it was found that 27.1% of
the crash were misclassified

34
00:01:47,670 --> 00:01:52,200
in terms of whether they
were curve-related or not.

35
00:01:52,200 --> 00:01:55,590
- They developed an automated
method to check the data

36
00:01:55,590 --> 00:01:57,960
and they used curve data
that was provided by TBP,

37
00:01:57,960 --> 00:02:00,270
which is a new dataset
that we've just recently

38
00:02:00,270 --> 00:02:02,190
made available over the last couple years.

39
00:02:02,190 --> 00:02:03,990
And so they went through
and they did the comparison.

40
00:02:03,990 --> 00:02:06,960
They took what they had
in the crash record system

41
00:02:06,960 --> 00:02:08,310
after understanding the data source.

42
00:02:08,310 --> 00:02:10,380
They took the GIS data that was available

43
00:02:10,380 --> 00:02:12,900
through our GS Online data services

44
00:02:12,900 --> 00:02:14,430
and they compared the two.

45
00:02:14,430 --> 00:02:18,168
- Using the automated procedure
that have been developed

46
00:02:18,168 --> 00:02:22,710
under this project will
allow TxDOT engineers

47
00:02:22,710 --> 00:02:27,360
to better identify curve-related crashes,

48
00:02:27,360 --> 00:02:32,360
use the portal TxDOT's safety
goal of zero fatalities.

49
00:02:33,030 --> 00:02:35,340
- Anytime you look at your
data through a different lens

50
00:02:35,340 --> 00:02:36,330
or compare it to something else,

51
00:02:36,330 --> 00:02:37,770
you find things that, you know,

52
00:02:37,770 --> 00:02:40,020
that you can improve and
that's what they did.

53
00:02:40,020 --> 00:02:41,250
- [Presenter] For more information

54
00:02:41,250 --> 00:02:44,190
and to find the publications
for this project,

55
00:02:44,190 --> 00:02:46,770
please visit the TxDOT Research Library

56
00:02:46,770 --> 00:02:48,483
at the link shown below.