This document analyzes GPS data collected from vehicles in Milan between April 1-7, 2007. It finds that:
1) Visitors' trips are significantly shorter in duration and distance than residents', but speeds are similar, indicating visitors park outside the city center and use public transit.
2) Traffic patterns differ on weekdays vs weekends, with lower average traffic each time slot on weekends. Working days see equal morning and afternoon traffic, while weekends have more evening traffic.
3) Top points of interest for visitors are near the city center on weekends, while residents' trips are uniformly distributed. Weather impacts traffic levels more than day of the week.
2. Name Description
user_id Unique id related to a car
timestamp
Date and time on the detection of the GPS device mounted on the
car
day_of_week Specifying the day of the week on the recordingade by GPS
time_slot Day parting based on the time recorded by the GPS device
lat Latitude of detection
lon Longitude of detection
GPS detection in Milan in the period between 1° April and April 7 2007
4. Days analyzed for residents: Sunday – Tuesday – Friday – Saturday
Days analyzed for visitors: Friday (from 1Pm) - Saturday
• Identify the most achieved attractions
(Point of Interest)
• Measuring the intensity of the traffic
according to the zone, the time of day
and to points of interest.
• Get insights thanks to the most
important route’s statistics (length,
duration, speed)
• Detect "behavioral" differences among
residents and visitors.
• Multidimensional analysis with
dimensions: weather - time - traffic
5. Visitors
Friday (from 1PM)
Saturday
(Split 2)
Residents
Sunday
Tuesday
Friday
Saturday
(Split 1)
Split Dataset into two parts
from Split 2 delete all users having user_id content in Split 1
7. The trend of high traffic flow in the city center and on high speed
highways, doesn't change with respect to residents.
VISITORS
8. Min Max AVG
Dom 0.367 368.483 40.56
Mart 0.3 367.983 44.731
Ven 0.25 325.25 43.681
Sab 0.55 290.467 43.098
0-30
30-60
60-90
90-120
120-150
150-180
180-210
210-240
240-270
270-300
300-330
330-360
360-390
1177
457
262
109
56 44 1 3 2 1 1 0 1
Distribuzione delle durate Domenica
56%22%
12%
5%
Composizione delle durate Domenica
0-30
30-60
60-90
90-120
120-150
150-180
180-210
Duration
• < 30 minutes: increase from Friday to Sunday, lower on Tuesday
• 30 – 60 minutes: decrease from Friday to Sunday, high on Tuesday
• >60 minutes: linear trend during the week
9. Speed
Min Max AVG
Dom 5.002 136.051 31.643
Mart 5.01 133.097 27.782
Ven 5.02 125.157 28.907
Sab 5.032 137.392 30.983
0-10
10-20
20-30
30-40
40-50
50-60
60-70
70-80
80-90
90-100
100-110
110-120
120-130
130-140
244
556
417
287
193
136
103
72
38 18 18 13 5 2
Distribuzione delle velocità Domenica
12%
26%
20%14%
9%
6% 5%
Composizione delle velocità Domenica
0-10
10-20
20-30
30-40
40-50
50-60
60-70
• No significant differences between working days speed and weekend days ones
• Probably why are there especialy urban detections?
10. Min Max AVG
Dom 0.047 152.287 16.193
Mart 0.059 139.366 15.719
Ven 0.029 126.646 15.28
Sab 0.108 125.017 15.823
Length
0-11
11-22
22-33
33-44
44-55
55-66
66-77
77-88
88-99
99-110
110-121
121-132
132-143
143-154
871
717
336
125
37 10 6 2 3 3 2 1 0 1
Distribuzione delle lunghezze Domenica
41%
34%
16%
6%
Composizione delle lunghezze Domenica
0-10
10-20
20-30
30-40
40-50
50-60
60-70
• Short trips (0-20 Km) have the opposite trend than long trips (>20 Km)
• Descend trend for short trips and ascend trend for long trips
11. VisitorsDuration
454
79
42
13 3 3
Distribuzione durate dei visitatori Venerdì
1395
227
113
47 26 10
Distribuzione delle dei visitatori Sabato
76%
13%
7%
2%
1%
1%
Composizione delle durate dei visitatori Venerdì
0-30
30-60
60-90
90-120
120-150
150-180
77%
12%
6%
3% 1%1%
Composizione delle durate dei visitatori Sabato
0-30
30-60
60-90
90-120
120-150
150-180
12. Residenti Giorno Visitatori
0 – 30: 54%
30 – 60: 21%
Venerdì
0 – 30: 76%
30 – 60: 13%
0 – 30: 55%
30 – 60: 19%
Sabato
0 – 30: 77%
30 – 60: 12%
Visitors short trips extremely greater than residents short trip
Visitors don’t use own car to move in Milano
13. Speed
41
102
88
64
41 45
29 24
33
43
35
24 21
4
Distribuzione delle velocità dei
visitatori Venerdì
7%
17%
15%
11%7%
8%
5%
4%
5%
7%
6%
4% 3%
Composizione delle velocità dei visitatori Venerdì
0-10
10-20
20-30
30-40
40-50
50-60
60-70
126
290
253
180
133
115
92 9010293
124116
66
28
10
Distribuzione delle velocità dei
visitatori Sabato
7%
16%
14%
10%
7%6%5%
5%
6%
5%
7%
6%
4%
Composizione delle velocità dei visitatori Sabato
0-10
10-20
20-30
30-40
40-50
50-60
60-70
14. Residenti Giorno Visitatori
0 – 30: 59%
0 – 50: 82%
+50: 18%
Venerdì
0 – 30: 39%
0 – 50: 57%
+50: 43%
0 – 30: 61%
0 – 50: 81%
+50: 19%
Sabato
0 – 30: 40%
0 – 50: 53%
+50: 47%
Visitors speed extremely lower than residents speed
Visitors park in Milano suburbs and use metro or TrenoNord
train company
15. Length
193 203
112
66
13 3 3 1
Distribuzione delle lunghezze dei
visitatori Venerdì
32%
34%
19%
11%
Composizione delle lunghezze dei visitatori
Venerdì
0-10
10-20
20-30
30-40
40-50
50-60
603 621
321
219
38
5 5 3 3
Distribuzione delle lunghezze dei
visitatori Sabato
33%
34%
18%
12%
Composizione delle lunghezze dei visitatori
Sabato
0-10
10-20
20-30
30-40
40-50
50-60
16. Residenti Giorno Visitatori
0 – 20: 73%
0 – 30: 89%
Venerdì
0 – 20: 66%
0 – 30: 85%
0 – 20: 70%
0 – 30: 85%
Sabato
0 – 20: 67%
0 – 30: 85%
On the opposite to duration and speed, length of trips in unchanged between
visitors and residents.
17. Time Label
00:00 – 05:59 Night
06:00 -08:59 Early Morning
09:00 – 11:59 Morning
12:00 – 13:59 Lunch
14:00 – 17:59 Afternoon
18:00 – 19:59 Early Evening
20:00 – 21:59 Evening
22:00 – 23:59 Late Evening
18. Time slots inhomogeneous? NORMALIZE!
Night
Early
Mornin
g
Mornin
g
Lunch
Afterno
on
Early
Evening
Dinner
Late
Evening
Domenica 35,5 92 106,3333 130,5 150,75 99,5 84,5 32,5
Martedì 119,1667 250 190 219 296,5 159 132,5 49,5
Venerdì 105,6667 236 189,6667 224,5 239 161,5 118 68,5
Sabato 42 116,6667 117 136,5 141 84 98,5 57,5
0
50
100
150
200
250
300
350
Corsemedieorarieperfasce
Intesità di traffico per fasce orarie
• Average traffic on weekend in each time slot is
always lower than the working days
• Working Cycle: (early morning + morning) is
almost equal to (afternoon + early evening)
19. Visitors
9%
17%
16%
13%
25%
8%
8% 4%
Composizione del traffico visitatori Sabato
Night
Early Morning
Morning
Lunch
Afternoon
Early Evening
Dinner
Time Slot
# trajectories
visitors
# trajectories
residents
Night 164 252
Early
Morning
307 350
Morning 292 351
Lunch 227 273
Afternoon 459 564
Early
Evening
143 168
Dinner 146 197
Late
Evening
80 115
• Friday not considered
• Decrease of the average trajectories in
visitors
• Trend unchanged
20. We partition the detection area in this way:
• Center: into the yellow perimeter
• Macroareas neighboring: between the yellow perimeter and mains roads
• Suburbs and airports: Externaly geographic area and intersects red perimeter
Macrozones flows analysis
21. O/D Matrix zones
The obtained results are equivalent, this highlights similar trend between working
days and holidays ones.
22. PO flows analysis
Po Rif. Po Rif.
Piazza Duomo 1 Zona Navigli 14
Galleria Vittorio Emanuele 1 Carcere San Vittore 15
Palazzo Reale 1 Basilica Sant'Ambrogio 16
Teatro Scala 1 Università Cattolica 17
Piazza Mercanti 1 Biblioteca Ambrosiana 18
Castello sforzesco 2 Politecnico 19
Parco Sempione 3 Porta Ticinese 20
Pinacoteca Brera 4 Corso Bueno s Aires 21
Orto Botanico 5 Basilica San lorenzo 22
Stadio Giuseppe Meazza 6 Porta Romana 23
Cimitero monumentale 8 Grattacielo Pirelli 24
Università Bocconi 9 Bosco Verticale 25
Stazione Ferroviaria Centrale 10 Museo Nazionale della Scienza 26
Aereoporto Linate 11 Chiesa San Sepolcro 27
Direzione Aereoporto Malpensa 12 Galleria Arte Moderna 28
Ospedale 13
23. PO Numero di arrivi
Direzione Malpensa 158
Aeroporto Linate 49
Piazza Duomo 9
Ospedale 4
Zona Navigli 3
Corso Buenos Aires 3
Bosco Verticale 2
Stazione Centrale 1
Porta Ticinese 1
Museo Arte Moderna 1
San Vittore 1
Not all PO present in the OD Matrix,
probabily for the difficulty to park in
centers.
24. To confirm this trend there is the
completely cluster end absence in
the historic centre.
27. Traffic flow analysis towards the airport
0
2
4
6
8
10
12
14
16
18
05:00-05:59
06:00-06:59
07:00-07:59
08:00-08:59
09:00-09:59
15:00-15:59
16:00-16:59
17:00-17:59
18:00-18:59
19:00-19:59
Partenze da Linate
Martedì start
Venerdì start
AVG week start
0
2
4
6
8
10
12
14
16
18
05:00-05:59
06:00-06:59
07:00-07:59
08:00-08:59
09:00-09:59
15:00-15:59
16:00-16:59
17:00-17:59
18:00-18:59
19:00-19:59
Arrivi a Linate
Martedì end
Venerdì end
AVG week end
28. • Despite the centre is very busy, we have mainly passing trajectories
• Denied expectations on the consequences of traffic decrease over the weekend
• average speed don't increase
• average length of shorts routes increase
• Visitors short trips extremely greater than residents short trip
• Visitors don’t use own car to move in Milano
• Visitors speed extremely lower than residents speed
• Visitors park in Milano suburbs and use metro or TrenoNord train
company
• On the opposite to duration and speed, length of trips in unchanged between
visitors and residents
• Average traffic on weekend in each time slots is always lower than the working
days
• The points of interest are not nearly ever reached by car
• The weather affects especially the highway and the roads to Linate Airport.
These trends are valid for both residents and visitors
37. 10%
13%
15%
12%
29%
10%
8% 3%
Composizione del traffico residenti Domenica
Night
Early Morning
Morning
Lunch
Afternoon
Early Evening
Dinner
Fascia Numero traiettorie
Night 213
Early
Morning
276
Morning 319
Lunch 261
Afternoon 603
Early
Evening
199
Dinner 169
Late Evening 65
17%
17%
13%
10%
27%
8% 6% 2%
Composizione del traffico residenti Martedì
Night
Early Morning
Morning
Lunch
Afternoon
Early Evening
Dinner
Fascia Numero traiettorie
Night 715
Early
Morning
750
Morning 570
Lunch 438
Afternoon 1186
Early Evening 318
Dinner 265
Late Evening 99
38. 16%
18%
14%11%
24%
8% 6%
3%
Composizione del traffico residenti Venerdì
Night
Early Morning
Morning
Lunch
Afternoon
Early Evening
Dinner
Fascia Numero traiettorie
Night 634
Early
Morning
708
Morning 569
Lunch 449
Afternoon 956
Early Evening 323
Dinner 236
Late Evening 137
11%
15%
16%
12%
25%
7%
9%
5%
Composizione del traffico residenti Sabato
Night
Early Morning
Morning
Lunch
Afternoon
Early Evening
Dinner
Fascia Numero traiettorie
Night 252
Early Morning 350
Morning 351
Lunch 273
Afternoon 564
Early Evening 168
Dinner 197
Late Evening 115