1
00:00:00,000 --> 00:00:16,560
<i>hacc preroll music</i>

2
00:00:16,560 --> 00:00:21,340
Herald: And a lovely welcome back to the
haccs stage on the third day this

3
00:00:21,340 --> 00:00:26,650
Congress, we are here with a talk on "A
few quantitive thoughts on parking in

4
00:00:26,650 --> 00:00:35,760
Marburg" by Martin L. He's interested in
data analytics now and infrastructure and

5
00:00:35,760 --> 00:00:41,059
traffic in general. And because of that,
he started scraping publicly available

6
00:00:41,059 --> 00:00:46,510
parking data in Marburg and just went on
and analyzed it and found a lot of

7
00:00:46,510 --> 00:00:52,360
interesting things which he is going to
present in this talk to you right now. In

8
00:00:52,360 --> 00:00:59,260
case you didn't know, there is IRC client
on the live.hacc.media where you can ask

9
00:00:59,260 --> 00:01:04,229
questions later or with the #rC3hacc tag
on Twitter.

10
00:01:04,229 --> 00:01:09,650
Martin Lellep: Welcome to my talk "A few
quantitative thoughts on parking in

11
00:01:09,650 --> 00:01:15,670
Marburg". I am delighted to speak here on
this Congress because I love the yearly

12
00:01:15,670 --> 00:01:20,130
conferences. Also, thank you to the
organizing team for making all this

13
00:01:20,130 --> 00:01:25,990
possible. You do an absolutely fabulous
job. Now, the first question that you

14
00:01:25,990 --> 00:01:32,540
should ask is: why? The following is a
purely hobby project question, I came up

15
00:01:32,540 --> 00:01:37,140
with a question because transportation is
important, but unfortunately, it's also

16
00:01:37,140 --> 00:01:43,071
difficult. The most popular vehicles these
days are cars and hence the question, how

17
00:01:43,071 --> 00:01:50,050
do people park in Marburg? Who am I? My
name is Martin, and I analyze publicly

18
00:01:50,050 --> 00:01:58,550
available data. I live close to Marburg,
therefore the parking in Marburg. Now, a

19
00:01:58,550 --> 00:02:04,700
little bit of background regarding
Marburg, it's a small picturesque, vibrant

20
00:02:04,700 --> 00:02:10,119
university town. There are a few highlights,
such as the castle, the old town and the

21
00:02:10,119 --> 00:02:16,560
river, just to name a few. It has around
80,000 residents and a somewhat dense core

22
00:02:16,560 --> 00:02:22,450
around the old town. You can see a few
pictures here of the castle, the old town

23
00:02:22,450 --> 00:02:29,670
and the river, respectively. Now, at this
point, I would like to give my props to

24
00:02:29,670 --> 00:02:34,170
David Kriesel because all this work was
inspired by his amazing data science

25
00:02:34,170 --> 00:02:39,330
talks. You can find them on YouTube. And I
absolutely encourage you to look for the

26
00:02:39,330 --> 00:02:49,210
Bahnmining, Spiegelmining and the Xerox
story talks. OK, so if you have questions,

27
00:02:49,210 --> 00:02:53,840
then please ask, I will be there live
during the Q&A of this conference and also

28
00:02:53,840 --> 00:03:01,510
you can send me an email with whatever you
like, essentially. OK, so first of all, I

29
00:03:01,510 --> 00:03:06,530
would like to give a quick introduction to
the data source. Now, the data, the

30
00:03:06,530 --> 00:03:13,890
parking data from Marburg is publicly,
well it's published live on a system that

31
00:03:13,890 --> 00:03:18,239
is implemented by the city, by the city
council, I believe . It's called

32
00:03:18,239 --> 00:03:26,230
Parkleitsystem Marburg or PLS for now, and
it publishes the data such as the parking

33
00:03:26,230 --> 00:03:31,599
decks, the number of free parking spots
and the location. The address here is

34
00:03:31,599 --> 00:03:39,069
pls.marburg.de. And let's see how it
looks. Yeah, so obviously it's still

35
00:03:39,069 --> 00:03:44,320
online and you can see here the parking
deck names listed, the number of free

36
00:03:44,320 --> 00:03:52,069
parking spots. Color coded is if it is
rather full or if it's rather empty, you

37
00:03:52,069 --> 00:03:58,390
can see here all of them are in the green.
The green color coding here, it's

38
00:03:58,390 --> 00:04:03,120
because it's probably close to Christmas.
Nobody wants to really park in the city.

39
00:04:03,120 --> 00:04:09,050
And the only one that's this one here, the
Marktdreieck Parkdeck that it has some

40
00:04:09,050 --> 00:04:15,650
load to it. Then also there's a button
called route. So whenever you click on the

41
00:04:15,650 --> 00:04:22,689
on this button, say we we pick the
Erlenring-Center button, we are redirected

42
00:04:22,689 --> 00:04:31,639
to Google Maps and we can see here the
location of this parking deck, for

43
00:04:31,639 --> 00:04:37,229
example. Let's go back. Last but not
least, there's also the maximum vehicle

44
00:04:37,229 --> 00:04:44,039
allowance and of course, the time stamp of
the data. OK, back to the presentation

45
00:04:44,039 --> 00:04:49,039
now. This is a very simple website, so of
course it's easy to scrape and that's what

46
00:04:49,039 --> 00:04:58,229
I did. Regarding the scraper, I used a
Linux computer and a docker container. And

47
00:04:58,229 --> 00:05:05,740
this scraper, you can see a small sketch
here to the left, it simply visits the

48
00:05:05,740 --> 00:05:11,360
website every 3 minutes inside the docker
container and writes the data into I

49
00:05:11,360 --> 00:05:17,509
believe it was csv files, which are
subsequently used for the data analysis.

50
00:05:17,509 --> 00:05:24,699
All of it, the scraper and the analysis
scripts are written in Python. OK, the

51
00:05:24,699 --> 00:05:33,219
data format is pretty simple, it's
processed internally with data frames,

52
00:05:33,219 --> 00:05:37,370
with the package panda. Everybody who
knows Python probably knows panda, anyway.

53
00:05:37,370 --> 00:05:42,619
It's the data format is as follows. The
row corresponds to the time. The column

54
00:05:42,619 --> 00:05:47,460
corresponds to the specific parking deck,
and the cell corresponds to the number of

55
00:05:47,460 --> 00:05:53,759
free parking spots at that time of that
parking deck. Now, in order to make the

56
00:05:53,759 --> 00:05:59,629
numbers a bit more usable, I transformed
the number of free parking spots to the

57
00:05:59,629 --> 00:06:05,550
number of used parking spots by
subtracting it from the maximum along the

58
00:06:05,550 --> 00:06:13,889
time. OK, now the intro is just to get
used to the data, we'd like to take a look

59
00:06:13,889 --> 00:06:19,529
at the locations of the of the park houses
or the park decks. This is a screenshot.

60
00:06:19,529 --> 00:06:26,599
There's an interactive version. Let me
open it here. It's a interactive map. You

61
00:06:26,599 --> 00:06:33,500
can see two types of markers, the first
one red, the second one green, and that's

62
00:06:33,500 --> 00:06:39,110
because the red ones are the ones that are
given, well they are encoded in the links

63
00:06:39,110 --> 00:06:45,809
of the PLS system, and they
are actually wrong. So when you click on

64
00:06:45,809 --> 00:06:54,460
the for instance. Erlenring-Center parking
deck that I've done before, the location,

65
00:06:54,460 --> 00:06:59,469
longitude and latitude are actually
incorrect and, um, Google Maps corrected

66
00:06:59,469 --> 00:07:04,409
on the fly. And therefore, I have shown
here the ones given on the website that

67
00:07:04,409 --> 00:07:11,259
are incorrect in red and the ones shown
that are correct. So you can safely focus

68
00:07:11,259 --> 00:07:17,069
only on the green ones. Um, a quick
overview here is the train station region,

69
00:07:17,069 --> 00:07:22,339
there are two. And then they are scattered
around the city. Um, sometimes there are

70
00:07:22,339 --> 00:07:30,159
two parking decks very close by, for
instance, these two and these two. And

71
00:07:30,159 --> 00:07:34,180
that's because it's essentially one
parking deck with two parking sections

72
00:07:34,180 --> 00:07:40,460
typically inside the building and on top
of the building. OK, let's go back to the

73
00:07:40,460 --> 00:07:49,089
presentation. With that in place, we or we
take a look at the joined data, meaning I

74
00:07:49,089 --> 00:07:55,639
accumulate the number of used parking
spots across all the parking decks. You

75
00:07:55,639 --> 00:08:00,110
can see that here now, so it's a quite
comprehensive picture, I started data

76
00:08:00,110 --> 00:08:06,479
scraping in August 2019 and stopped it at
the end of February 2020.

77
00:08:06,479 --> 00:08:13,179
This data here is a different resample
frequency of the original and raw data. I

78
00:08:13,179 --> 00:08:17,370
started with a resample of one hour. So
just a reminder, it's the true frequency

79
00:08:17,370 --> 00:08:23,279
is three minutes. Again, I resampled here
into one hour. It's not very easy to

80
00:08:23,279 --> 00:08:27,839
understand on that scale here. Then to one
day it's the orange now and lastly on one

81
00:08:27,839 --> 00:08:33,490
week and we can learn different things
from it. So in particular, the orange

82
00:08:33,490 --> 00:08:38,760
curve of one day shows that there might be
some periodicity in the signal. And the

83
00:08:38,760 --> 00:08:43,870
green one shows that there are times or
weeks that are particularly... where

84
00:08:43,870 --> 00:08:47,640
there's particularly little parking
demand, for instance, here around

85
00:08:47,640 --> 00:08:56,490
Christmas 2019. OK, so again, from the
orange signal, you can see that there's

86
00:08:56,490 --> 00:09:01,570
probably some periodicity, and in order to
quantify that, I plotted the or computed

87
00:09:01,570 --> 00:09:06,660
the auto correlation function. The auto
correlation function essentially takes a

88
00:09:06,660 --> 00:09:11,570
time signal and computes the overlap
between the time signal and the same

89
00:09:11,570 --> 00:09:17,740
signal shifted by some time and whenever
there's a large overlap. That points

90
00:09:17,740 --> 00:09:23,710
towards the periodicity, and here we can
see that the periodicity maximum or the

91
00:09:23,710 --> 00:09:28,440
auto correlation maximum, the first one
corresponds to one week and therefore the

92
00:09:28,440 --> 00:09:33,690
periodicity can be safely assumed to be at
seven days. Of course, when there's

93
00:09:33,690 --> 00:09:41,060
periodicity and a signal at seven days,
for instance, there's also periodicity. In

94
00:09:41,060 --> 00:09:44,680
14 days and in 21 days, but the
correlation coefficients, they decay

95
00:09:44,680 --> 00:09:53,470
typically. OK, now we have the periodicity
with respect to days in place. Now let's

96
00:09:53,470 --> 00:09:59,010
take a look at the day and hour demand.
And for that, I computed a two dimensional

97
00:09:59,010 --> 00:10:05,950
histogram with the day Monday to Sunday on
the one axis and the other axis

98
00:10:05,950 --> 00:10:11,780
corresponds to the hour. And here we can
clearly see that the majority of the

99
00:10:11,780 --> 00:10:17,820
parking demand is around the noon hour. So
starting from 11 to to approximately,

100
00:10:17,820 --> 00:10:26,270
let's say, 5 p.m. or so. Interestingly.
That was a point where I was surprised is

101
00:10:26,270 --> 00:10:30,740
that Sunday's is a day where there's
little parking demand in Marburg, I

102
00:10:30,740 --> 00:10:35,720
wouldv'e guesstimated that Sunday when
everybody has spare time, they typically

103
00:10:35,720 --> 00:10:40,270
rush into the city. But that's obviously
not the case. Another interesting fact is

104
00:10:40,270 --> 00:10:44,840
that Monday morning seemed to be very
difficult to get up because you can see

105
00:10:44,840 --> 00:10:55,340
the parking demand is smaller than on on
other mornings. OK, now, after that, I

106
00:10:55,340 --> 00:11:01,490
come to the separated... separate and
analysis where I take a look at the

107
00:11:01,490 --> 00:11:07,520
individual parking decks. So first of all,
again, the times series, it's it's a bit

108
00:11:07,520 --> 00:11:11,540
dense and it's very hard to see. So there
are a few things to learn from the

109
00:11:11,540 --> 00:11:16,410
picture. So first of all, the green
signal that corresponds to the Erlenring-

110
00:11:16,410 --> 00:11:22,320
Center. Reminder, I just opened it. In the
very beginning of this talk seems to be

111
00:11:22,320 --> 00:11:32,500
the dominant one, then there are quite a
few data gaps. So take for instance. Well,

112
00:11:32,500 --> 00:11:37,370
it's very apparent here for the violet
one, the Furthstraße-Parkdeck, this one

113
00:11:37,370 --> 00:11:44,530
here. And that's an extreme case. It had
obviously some kind of problem. It was

114
00:11:44,530 --> 00:11:49,280
open for some time and then closed for
some other times. Typically, park houses

115
00:11:49,280 --> 00:11:54,280
or parking decks are either open 24/7, but
there are also quite a few that are that

116
00:11:54,280 --> 00:12:07,030
close overnight. OK, next I was interested
in the statistics of parking demand for

117
00:12:07,030 --> 00:12:13,920
individual parking decks, so I
concentrated only on, say, one parking

118
00:12:13,920 --> 00:12:20,680
deck and computed the histograms of the
used parking spots also, depending on the

119
00:12:20,680 --> 00:12:27,530
time. Let's focus here on the Oberstadt,
it's the old town and you can see that the

120
00:12:27,530 --> 00:12:37,130
overall parking demand peaks at around,
let's say, maybe 20 used parking spots, so

121
00:12:37,130 --> 00:12:42,470
that's the average, but that's not for all
times when we make that statement,

122
00:12:42,470 --> 00:12:45,600
depending on the time, for instance, the
morning we can see that's approximately

123
00:12:45,600 --> 00:12:51,630
the same. But when we go towards noon, we
can see that the number of parking spots

124
00:12:51,630 --> 00:12:58,680
or used parking spots increases. There are
even a few times when it's at the maximum

125
00:12:58,680 --> 00:13:05,500
around noon. Now, when we go towards later
hours, the maximum shifts towards smaller

126
00:13:05,500 --> 00:13:11,680
values again. Now, this this behavior of
the maximum shifting, so clearly,

127
00:13:11,680 --> 00:13:17,150
depending on the hour, is not not apparent
for all the parking decks. For instance,

128
00:13:17,150 --> 00:13:24,010
the Parkdreieck here ... Marktdreieck,
sorry, that doesn't show the signal as

129
00:13:24,010 --> 00:13:32,890
clear as the Oberstadt one. OK, from this
all now we can quantify also the, I call

130
00:13:32,890 --> 00:13:37,660
it integral parking demand, simply it's
the the number of parking spots that have

131
00:13:37,660 --> 00:13:44,760
been provided per parking deck. Now the
picture here, it's normalized to the

132
00:13:44,760 --> 00:13:49,700
maximum and one can see from this picture
here very easily that the Erlenring-

133
00:13:49,700 --> 00:13:54,200
Center, as we've estimated or guessed
previously already is the one that's

134
00:13:54,200 --> 00:14:01,720
dominating the whole city. It's providing
the most parking spots by a large margin,

135
00:14:01,720 --> 00:14:07,690
actually. The next one is the Lahn-Center
and then maybe the Oberstadt and the other

136
00:14:07,690 --> 00:14:12,750
ones follow after these. Another
interesting point here is that the

137
00:14:12,750 --> 00:14:20,610
proportion of parking spots provided on
weekends differs for the different parking

138
00:14:20,610 --> 00:14:25,300
decks. For instance, here you can see this
one here is quite a big portion, the

139
00:14:25,300 --> 00:14:29,740
Erlenring-Center, also on weekends.
Contrary, the Marktdreieck-Parkdeck has

140
00:14:29,740 --> 00:14:38,420
only a very small portion of, um, of
parking spots provided on weekends. It

141
00:14:38,420 --> 00:14:43,520
might be interesting to know that this
particular parking station is ... it's the

142
00:14:43,520 --> 00:14:47,880
one that is used if you want to go to a
doctor, because it's very close. So many

143
00:14:47,880 --> 00:14:51,090
doctors are not open on Sundays, on
Saturdays, and therefore probably the

144
00:14:51,090 --> 00:14:56,130
parking demand is quite low. Now, there's
a temporal version also where I rendered a

145
00:14:56,130 --> 00:15:02,230
small video that I'm opening now, and you
can see essentially the same as in the

146
00:15:02,230 --> 00:15:07,700
previous graph, but against time. Again,
it's very apparent that there's a

147
00:15:07,700 --> 00:15:15,590
periodicity and here my scraper crashed
and it's back in business again, and I

148
00:15:15,590 --> 00:15:23,100
found it interesting to see that there are
parking decks that have cars... well that

149
00:15:23,100 --> 00:15:28,040
host cars, even at night, for instance,
here the Erlenring-Center again in the

150
00:15:28,040 --> 00:15:33,500
Lahn-Center, the ones that are the largest
one, they offer parking also overnight.

151
00:15:33,500 --> 00:15:42,360
And there are some cars in there,
probably. OK, let's close that again. Now,

152
00:15:42,360 --> 00:15:51,000
I come lastly to the prediction part now.
The goal here is to measure the parking

153
00:15:51,000 --> 00:15:56,339
demand through the parking decks, but then
to interpolate between the parking decks,

154
00:15:56,339 --> 00:16:02,170
so I would like... so I have ...say the
Oberstadt the old town and the, I don't

155
00:16:02,170 --> 00:16:05,540
know, the Erlenring, which was the largest
one. I would like to know what's the

156
00:16:05,540 --> 00:16:12,070
parking demand in between, for instance.
For doing so, I use a spatial fit and I

157
00:16:12,070 --> 00:16:16,420
use a machine learning model for that, in
order to do that spatial fit. It is now,

158
00:16:16,420 --> 00:16:22,140
in this particular case, a non parametric
model called Gaussian Process Regression.

159
00:16:22,140 --> 00:16:27,170
And the nice thing about that is that it
also returns the uncertainty. Because say,

160
00:16:27,170 --> 00:16:31,400
for instance, you would like to use these
model, machine learning predictions to

161
00:16:31,400 --> 00:16:37,670
say, build some kind of parking deck or to
get rid of one. All these operations, all

162
00:16:37,670 --> 00:16:41,890
these derived actions would be very
expensive. So you would like to know if

163
00:16:41,890 --> 00:16:45,800
the uncertainty is large or small for
whatever the machine learning model

164
00:16:45,800 --> 00:16:51,450
predicts. Just for the math oriented
people. If you're interested in that

165
00:16:51,450 --> 00:16:57,140
model, definitely take a look at the, I
would call it, Gaussian process bible by

166
00:16:57,140 --> 00:17:05,890
Rasmussen. It's amazing to read. Yeah,
there are two, um, evaluations now, I did.

167
00:17:05,890 --> 00:17:09,549
The first one is based on the whole data
set, so there's no spatial or..sorry...

168
00:17:09,549 --> 00:17:16,710
there's no temporal resolution. And what I
do, I did well, I rrendered a video and I

169
00:17:16,710 --> 00:17:22,920
would like to explain you the outcome of
that while it is running. The top picture

170
00:17:22,920 --> 00:17:26,929
here shows you the prediction by the
machine learning model. And the the bottom

171
00:17:26,929 --> 00:17:35,250
picture shows you the uncertainty. The
training data, meaning the parking decks,

172
00:17:35,250 --> 00:17:41,169
is denoted by the black points. Now, first
of all, the uncertainty, you can see that

173
00:17:41,169 --> 00:17:46,220
wherever there is training data, the
uncertainty goes down. So the model is,

174
00:17:46,220 --> 00:17:50,619
um, certain about its prediction that
because, well, there's training data and

175
00:17:50,619 --> 00:17:57,690
in between the uncertainty rises again.
Now the prediction, you can see some small

176
00:17:57,690 --> 00:18:05,639
hill. It's exactly the Erlenring-Center,
which was the largest one. Now, what is

177
00:18:05,639 --> 00:18:10,720
shown in the video is it's rotating. You
can see the coordinates of Marburg on the

178
00:18:10,720 --> 00:18:16,350
on the plane, on the bottom plane. And at
some point, the view rotates upwards and

179
00:18:16,350 --> 00:18:21,140
gives you a top down perspective with a
corresponding color bars or corresponding

180
00:18:21,140 --> 00:18:28,850
color map. So, again, here's the the
maximum, the Erlenring-Center. And I did

181
00:18:28,850 --> 00:18:37,149
that because next we would like to finally
measure the parking demand between

182
00:18:37,149 --> 00:18:44,570
stations. OK, there's another small video
again, and now we start right from the top

183
00:18:44,570 --> 00:18:48,610
down, color coded view and again, the
black points are the... is the training

184
00:18:48,610 --> 00:18:53,119
data, but now the red points are, is kind
of test data, meaning positions in

185
00:18:53,119 --> 00:18:59,740
between. I concentrated now on the Mensa
because I have a special relation with the

186
00:18:59,740 --> 00:19:04,730
Mensa, the physics department, the
university library, the train station and

187
00:19:04,730 --> 00:19:09,789
the cinema. And just to demonstrate from
this spatial fit, we can derive the

188
00:19:09,789 --> 00:19:14,740
parking demand at these positions also.
Here, this yellow pike, it's the

189
00:19:14,740 --> 00:19:21,499
Erlenring-Center again. Now, that's only a
qualitative result, of course, I don't

190
00:19:21,499 --> 00:19:25,169
want to derive any quantitative at this
point, it's just a proof of concept that

191
00:19:25,169 --> 00:19:32,400
it is possible to derive something like
that from the publicly available data.

192
00:19:32,400 --> 00:19:36,500
Now I forgot to mention the beginning that
there's a bonus and I would like to come

193
00:19:36,500 --> 00:19:43,529
to the bonus now. It is about the Corona
crisis or pandemic, of course. What I did

194
00:19:43,529 --> 00:19:48,019
is, the initial data acquisition phase,
here in black, that's the whole talk was

195
00:19:48,019 --> 00:19:54,309
about that black portion here. I stopped
it at around the end of February and I

196
00:19:54,309 --> 00:20:00,119
restarted the whole data acquisition
process now again at in approximately

197
00:20:00,119 --> 00:20:05,590
April. Just to capture something from the
Corona crisis as well. And you can see

198
00:20:05,590 --> 00:20:12,049
here again, the time series. I think the
most interesting bit about it and the most

199
00:20:12,049 --> 00:20:16,739
comprehensive bit is the the mean . You
can see the the mean across the whole time

200
00:20:16,739 --> 00:20:20,820
denoted by this dashed line. And you can
see that the mean is smaller. So during

201
00:20:20,820 --> 00:20:27,379
the Corona pandemic fewer people parked in
Marburg, which is reasonable, I would say.

202
00:20:27,379 --> 00:20:34,690
But there are also times where the number
of parking spots decreased significantly.

203
00:20:34,690 --> 00:20:40,249
So for instance, right when the Corona
crisis started in April and now the second

204
00:20:40,249 --> 00:20:46,340
wave in October, November, December, it is
visible that the parking demand decreased

205
00:20:46,340 --> 00:20:53,220
a lot. And I went one step further and
wanted to know the the differences between

206
00:20:53,220 --> 00:20:59,429
pre Corona and during Corona also for each
of the parking decks, that's what I did

207
00:20:59,429 --> 00:21:04,730
here. It's now not the normalized parking
demand, but the absolute parking demand.

208
00:21:04,730 --> 00:21:10,539
So now we can see also the absolute
numbers, the black black bars you've seen

209
00:21:10,539 --> 00:21:16,440
previously already. Now the red bars is
for the during the Corona crisis. And then

210
00:21:16,440 --> 00:21:22,340
I defined these, the first wave and the
second wave as serious corona times. So I

211
00:21:22,340 --> 00:21:25,831
also plotted a third bar... set of bars
here. And it's interesting to see that

212
00:21:25,831 --> 00:21:32,730
while most of the parking decks, of
course, suffered in terms of providing

213
00:21:32,730 --> 00:21:39,630
parking demands or most of them provided
fewer parking decks, parking spots. But

214
00:21:39,630 --> 00:21:44,119
there are a few, like, for instance, the
Marktdreieck-Parkdeck here that, well,

215
00:21:44,119 --> 00:21:49,710
almost increased. We can see during the
corona in general it increased a bit. And

216
00:21:49,710 --> 00:21:54,190
then during the heavy corona, it increased
even more. And as I mentioned before, this

217
00:21:54,190 --> 00:22:00,329
is the parking deck that corresponds to,
yeah, a whole collection of doctors. So. I

218
00:22:00,329 --> 00:22:07,220
derive that well during Corona times the
parking demand in front of doctors even

219
00:22:07,220 --> 00:22:13,369
increased a tiny bit. Yeah, with that, I
would like to come to my conclusions.

220
00:22:13,369 --> 00:22:18,100
Thank you for sticking with me until now.
So I scraped publicly available data here

221
00:22:18,100 --> 00:22:26,840
with a small scraper set up. I analyzed
it, for instance, for day and hour

222
00:22:26,840 --> 00:22:30,889
patterns. And last but not least, did some
machine learning in order to quantify the

223
00:22:30,889 --> 00:22:38,779
demand in between the stations, there is
an accompanying blog article also. You can

224
00:22:38,779 --> 00:22:42,169
find it down here, there all the figures
in higher resolution and you can play

225
00:22:42,169 --> 00:22:47,519
around with an interactive map also, if
you like. Um, and to finally now conclude

226
00:22:47,519 --> 00:22:51,419
the presentation. I would like to hear
from you what you think about this

227
00:22:51,419 --> 00:22:57,460
analysis. I'd like to improve with these
kind of mini studies. And therefore, I

228
00:22:57,460 --> 00:23:03,059
would be very interested in your critique
regarding the content, the presentation

229
00:23:03,059 --> 00:23:08,019
and general content... comments. Again,
you can email me to this email address

230
00:23:08,019 --> 00:23:15,919
here, or alternatively, I set up a Google,
um, Google form. So the Google forms

231
00:23:15,919 --> 00:23:20,509
document that exactly comprised of these
questions, and you can simply type them in

232
00:23:20,509 --> 00:23:24,919
if you're interested. Thank you very much.

233
00:23:24,919 --> 00:23:29,559
Herald: All right, first of all, thank you
for this amazing talk, I have a few

234
00:23:29,559 --> 00:23:33,940
questions what have been relayed to me and
I'm just going to ask them one after the

235
00:23:33,940 --> 00:23:39,070
other. And let's not waste any time and
start with the first one. Have you found

236
00:23:39,070 --> 00:23:46,889
parking decks that are usually heavily
overloaded or never completely used?

237
00:23:46,889 --> 00:23:56,029
Martin: Um so. Given that there are only
around what was it, 8 or 9 or 10 in the

238
00:23:56,029 --> 00:24:05,159
data set, honestly, I never looked for for
that question. So, um, short answers is:

239
00:24:05,159 --> 00:24:10,779
No. Long answer, yes, I could have or I
still could, I would say.

240
00:24:10,779 --> 00:24:16,980
H: OK. Have you tried prediction in time,
so guessing which parking decks will be

241
00:24:16,980 --> 00:24:21,700
exhausted soon?
M: No, no. So that's obviously it's

242
00:24:21,700 --> 00:24:26,750
like... it's... I would consider that
something like the predictive maintenance

243
00:24:26,750 --> 00:24:32,929
of traffic business kind of. It's
definitely a thing that people that have

244
00:24:32,929 --> 00:24:38,409
more time and more are willing to invest
more definitely should do and could do. I

245
00:24:38,409 --> 00:24:43,769
would say I mean, there's lots of lots of
additional data that might be of interest,

246
00:24:43,769 --> 00:24:49,369
like weather data. And, for instance, is
it a is it a public holiday, yes or no and

247
00:24:49,369 --> 00:24:55,580
all that kind of stuff. So, again, short
answer.: No. Long answer. Yes. Would be

248
00:24:55,580 --> 00:25:00,919
possible.
H: OK, so if anyone watching has the time

249
00:25:00,919 --> 00:25:05,710
or energy to do that, they could.
M: Absolutely. Yes.

250
00:25:05,710 --> 00:25:10,649
H: OK, and the last question I have right
now is, will the code or especially the

251
00:25:10,649 --> 00:25:16,509
scraping part be available publicly or
like in the GitHub or somewhere?

252
00:25:16,509 --> 00:25:24,580
H: Um, I could do that. So I was very I
was quite hesitant with it. So obviously

253
00:25:24,580 --> 00:25:29,840
publishing the data could be problematic.
I have no experience with it on the legal

254
00:25:29,840 --> 00:25:34,150
side. So I would probably not publish the
data, which is I mean, it's old data

255
00:25:34,150 --> 00:25:41,039
anyway. So and but then regarding the
code, I was just waiting if anybody's

256
00:25:41,039 --> 00:25:45,109
interested. So given that somebody stated
the interest, I would probably publish it.

257
00:25:45,109 --> 00:25:49,639
Yes.
H: OK, yeah I think that's it from the

258
00:25:49,639 --> 00:25:52,179
question side .
M: Hmhm.

259
00:25:52,179 --> 00:25:58,559
H: And they were all answered quite
nicely. And judging by that, I don't get

260
00:25:58,559 --> 00:26:04,809
any more questions right now. So, yeah, I
would conclude is talk. Maybe you can also

261
00:26:04,809 --> 00:26:08,369
like have a last word. From my side I'm
done here.

262
00:26:08,369 --> 00:26:14,779
M: Yes. So, um, well, thank you very much
for watching the talk. And I try to

263
00:26:14,779 --> 00:26:19,600
improve. I think I said it on the last
slide. If I'm right, let me know if you

264
00:26:19,600 --> 00:26:26,109
have any doubts or things to improve
essentially on. And then regarding maybe

265
00:26:26,109 --> 00:26:31,720
the last question of publishing it, I
believe that I put a link there to find my

266
00:26:31,720 --> 00:26:37,700
blog and I would probably just add another
blog post stating well there's an github

267
00:26:37,700 --> 00:26:42,240
repository. You can go there and just find
just find the code and stuff like that

268
00:26:42,240 --> 00:26:48,260
there. So if you're interested, just, you
know, find my website. My name is Martin

269
00:26:48,260 --> 00:26:55,690
Lellep. Um, and then you will in a few
days, I guess probably in 2021 only. So I

270
00:26:55,690 --> 00:27:00,080
won't be able to publish it in the next
two days. But then the code will be

271
00:27:00,080 --> 00:27:06,049
public. Yes.
H: OK, then. Have a great day. Great time

272
00:27:06,049 --> 00:27:08,740
at Congress and byebye.

273
00:27:08,740 --> 00:27:10,810
<i>postroll music</i>

274
00:27:10,810 --> 00:27:38,000
Subtitles created by c3subtitles.de
in the year 2021. Join, and help us!