Google knows quite a lot about where I’ve been, apparently.

I’ve been using Google Maps since around 2010.
And since around 2010, Google has been storing my location history on their servers.

I decided request a copy of my location history from Google Takeout to see what they know
(or at least what they want to tell us that they know).
I selected the Location History (as JSON) option, waited for my archive to get created and downloaded the files.
Then I wrote a Python script to extract the information and plot the location data onto a map.

And well… Oh boy.
Google knew a lot.

Map

Overview

Below is a map of my travels in the last 4 years.
Anything past 4 years worth of data - and my computer couldn’t handle displaying all the 864647 data points at once…

Doesn’t look like that much to be honest, it shows that I’ve been to different parts of Australia, the United States, and a handful of Asian and European countries.
But let’s take a closer look.

Map (Sydney)

If we zoom the map into just the Sydney area, we can get a pretty good overview of my travels.

Quite a heavy presence down south - perhaps I might live there or something?

Conclusion

It’s quite frightening actually to know that Google retains my location data since around 2010.

And mind you, the map images above only show 4 years worth of data.
There’s still another 5 or so other years unrevealed.

9 years worth of location history. MY location history.

But of course, I knew that they had this data.

The decision to use Google Maps.
The decision to use a smartphone with a GPS module.
The decision to press “I Accept” on the terms and conditions.

We all know that other people have our data.
We trade our privacy for convenience - and in some cases we’re okay with that, other times not so much.

What about you?
What does Google know about you?

Code

 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

import json
import simplekml

def iterateData(data):
  for point in data:
    ts = int(point["timestampMs"])
    lat = int(point["latitudeE7"]) / 10000000
    long = int(point["longitudeE7"]) / 10000000
    yield(ts, (lat, long))

print("Parsing...")

data = json.load(open("Location History.json", "r"))

data = data['locations']

print("There are %d data points" % len(data))

fCount = 0
i = 0

kml = None
pieces = []

for timestamp, (lat, long) in iterateData(data):
  if i % 190000 == 0 and i != 0 and False:
    kml = simplekml.Kml()
    kml.newlinestring(name="line", coords=pieces)
    kml.save(f"output{fCount}.kml")
    pieces = []
    fCount += 1
  pieces.append((long,lat))
  i += 1

if len(pieces) > 0:
  kml = simplekml.Kml()
  kml.newlinestring(name="line", coords=pieces)
  kml.save(f"output{fCount}.kml")