ISSS608 2016-17 T3 Anuthama Unusual Patterns

From Visual Analytics and Applications
Jump to navigation Jump to search

Bluebird.gif The Missing Case of the Rose-Crested Blue Pipit

Background

Data Prep

Q1

Q2

Q3

Conclusion

Feedback

 


Unusual Patterns


In this page I have discussed some unusual patterns that were seen in the cars movement. I have called out specific vehicles/vehicle type.

Pattern 1: Is there a pattern in the intrusion at Gates?

  • The dashboard below reveals a few data points about the intrusion. Firstly not all gates have been trespassed, it is only gates 3,5,6.
  • The intrusions happen from 2 AM -5 AM which is the time Rangers are not around.
  • The intrusions happen only on 2 days Tuesdays and Thursdays.
100%

Pattern 2:No interest in the Preserve- just trespassing

When I analysed the paths taken by some of the car IDs I found something that caught my attention. Some vehicles have merely entered and exited the preserve and have not gone to any place in particular inside the preserve.

  • 2 passages were popular for trespassing
Entrance 2 <--> 4
Entrance 0 <--> 3
  • 1580 Vehicles have gone through just entrances without visiting any parts of the preserve.
  • The dashboard that has been filtered based on path to show the usage pattern reveals that May sees a surge in trespassing. This is in not in alignment with to the overall peak cycle for the preserve which is June-Aug.
  • All car types except 2P Ranger vehicles have used this pathway for trespassing through the preserve.


100%
100%



















Others

Under usual conditions a trespasser will have 2 sensor detection - one at the entrance and one at the exit. However, I suspect that more vehicles might have trespassed through the preserve but with one or two more sensors detecting the car IDs along the way between the entrances. So I filter for car ids with 4 sensor readings.When 4 sensor readings were taken as the filter condition we see that there are more car ids that possible would have trespassed through the preserve as its a short cut to their destination. There were 3196 such vehicles.

100%


Pattern 3: Suspicious behavior 1

We observe a strange pattern among these vehicles

  • These 4 vehicles came to the preserve at night and and were back again the next morning. The first day they entered via a particular entrance gate and exited via the same entrance gate, which means they are not here to trespass, but need to visit the entrance before they arrive the next morning.
  • Next morning they are back at the same entrance and took a trip into the preserve.
  • The paths taken appear different, but they appear to have gone through Ranger-stop 0 and 2. This could be the destination. They might have some work there.
  • All were of the same car type - 4 axle (and above) truck which means the purpose of visit may require a heavy vehicle.
  • Interestingly all of them spent most time at entrance on the second morning. Too many similarities for it to be a coincidence. They might be there delivering some goods and probably the goods were being checked at the entrance.


100%

Pattern 4:Suspicious behavior 2

Car ID 20162904122951-717 exhibited a strange repeated pattern.It is a 2 axle car (or motorcycle) This car has:

  • Visited the preserve every Friday for 6 weeks in a row from March 4th –April 10th and has followed the exact same sequence.
  • Camped over the weekend at camping 0 and has entered the camp and left through the same path -[Entrance 3 --> General gate 7 --> Camping 0 --> Camping 0 --> General gate 7 --> Entrance 3]. This car has not explored any other parts of the preserve. Could the visit be for a specific purpose and not so much for sight seeing?


100%

The following dynamic image shows the path taken by car ID 20162904122951-717 over the 6 trips. The different colours denote the week at which the trip was made.

100%

Pattern 5:Suspicious behavior 3

Car ID 20154519024544-322 is the most frequent flyer who had made 16 trips. Naturally I was curious to see the path of this frequent flyer.There was a strange repeated pattern in this visitor too. The car has:

  • Visited the preserve every week from June 19 – Oct 2 for 16 weeks in a row.
  • Camped at camping 4 each time it visited.
  • With the exception of the first trip all other trips have the same path. However in some cases the car left on Sunday from the camp and in some cases it left on a Monday.

I drilled down further on the data and did an hour wise analysis which revealed that this car arrived around 2-3PM on a Friday and goes to camping 4 and left the camp site on Sunday night closer to midnight which is the reason the car appears to have left the camp on Sunday on certain days and Monday on other days

100%

The following dynamic image shows the path taken by car ID 20154519024544-322 over the 16 trips. The different colours denote the week at which the trip was made.

100%


Pattern 6:New Home ??

Someone has found a new home in this preserve. Car ID 20155705025759-63 has been inside the preserve for almost a year now and still has not left the preserve. We see a pattern where this car goes to each camping site and stays there for weeks together.

100%
100%