Sunday, January 1, 2012

Research: Egyptian Traffic Analyses and Statistics ( Data from Bey2ollak )

The research below answers the following questions
  1. What are the rush hours in Egypt
  2. What are the worst/best routes in Cairo in terms of traffic
  3. What is the traffic load across hours and in weekends

The Research

I managed to conduct this research on a 1 month span using a script which loads data from Bey2ollak website every 30 minutes, the extracted data was saved and analyzed to deduce the conclusion below


Context:

Data: from bey2ollak website
Interval: every 30 min
Research duration: from 2/12/2011 - 30/12/2011


Conclusion

- The worst route in Cairo ( in terms of traffic ) is 6 of October bridge - both sides

- Highest point of traffic density (all days) is @ 18:30 PM

- Highest traffic density Hour (all days) is 18:00 PM

- Lowest traffic density Hour (all days) is 6:00 AM

- Highest traffic density span is from 15:00 PM to 20:00 PM

- Lowest traffic density span is from 4:00 AM to 6:00 AM

- Top day ( in terms of traffic ) is Tuesday

- Lowest day ( in terms of traffic ) is Friday

- Lowest week day ( in terms of traffic ) is Wednesday



Charts

Overall Traffic Density Rating for Cairo Roads





Cairo Traffic Density - Hourly










Overall Cairo Traffic Density - Daily















Overall Cairo Traffic Density - Weekends









Overall Traffic Density - Detailed



















Just don't be on Me7war @ 6 PM :)


Technical Stuff

Let me know if you need the script used in this research


Data

http://bit.ly/rTlRNK 


FAQ 

1- What does the values in the data sheet mean

The data ranges from "0" to "5" where "0" is the lowest traffic density and "5" is highest traffic density, each number corresponds to one of the well known "bey2ollak" labels as shown below

No data:      0
7alawa:       1
lazeez:        2
mashy:        3
za7ma:        4
mafeesh 2amal:    5

The script accessed bey2ollak every 30 min and took a snapshot of each road's label at that moment and mapped it to a number from 0-5, this is how the data sheet was created

2- How is the traffic density index calculated

The traffic density index is the sum of all data values (0-5) across the research duration ( 1 month ) for a specific hour, road or day


Similar Researches

http://karim-ouda.blogspot.com/2011/11/research-facebook-usage-in-egypt.html

http://karim-ouda.blogspot.com/2011/11/research-twitter-usage-pattern-in-egypt.html


Copyright/License


Creative Commons License
Egyptian Traffic Analyses and Statistics research by Karim Ouda is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.


24 comments:

  1. the most useful one, well done (y)

    ReplyDelete
  2. Great job ya karim (y)

    ReplyDelete
  3. GREAT JOB.. KEEP IT UP .. finally someone doing it ..

    ReplyDelete
  4. Kudus! I've been thinking about doing this for a long time, and was also thinking about building a model to predict future traffic! :)

    Was the website's terms of use ok with scraping their data, though? What stopped me was that I couldn't find a public API for bey2ollak. Well done, and thanks for the analysis.

    Would you be willing to share the data, btw?

    ReplyDelete
  5. Great effort ... how did you manage to measure the traffic? road cameras, other counters? just curiosity not really doubting the results

    ReplyDelete
  6. SA-
    great job masha2 Allah keep going. I have a few comments if u allow me:
    1- the work is based on web site which doe not indicate the traffic distribution instead it models the tweets about traffic in Egypt. There should be high correlation in some places but there will be no correlation in other areas. Consider for example the traffic in a place like Dar ElSalam, it is very crowded but you will not find this on the site because most of people does there do use the site. One possible extension is to predict the traffic given the number of tweets in certain place.
    2- the analysis of highest/lowest traffic in working days seems not accurate since I think the difference is not significant and no clear reason why is this. A possible step to validate your data is perform 2 tailed paired t-test.
    3- Other possible and nice extension would be the addition of a time interval without schools running t study the effect of schools on traffic.
    Generally, I like the idea and results very much and please keep going.

    Thank you,
    Ali ElKahky

    ReplyDelete
  7. I'm a BI Technology Solution proffessional in Microsoft... and I LUV IT my Friend, you made my day

    ReplyDelete
  8. SA
    Thanks for your effort really impressed
    excellent Idea, simple, automated,.. etc

    ReplyDelete
  9. Dear Karim,

    Thanks a lot for your effort. I have two notes.

    1-What is the definition of the measurement unit?
    2-A useful way to analyze the huge amount of data:
    2.1- Do the analysis of the ring road together.
    2.2-Do the analysis of the exit and entry to Cairo.
    2.3-Do the analysis for inside Cairo alone.
    2.4 Do a plot for these charts on Googlemaps.
    2.5-Suggest alternative if you can.

    These are only idea however, this is great work and bright idea from you. Welldone

    ReplyDelete
  10. Great job I am fully agree with Ali ElKahky comment specially with t-test to check data accuracy and the correlation with schools time in addition to Gov employee working time and the relation if possible with buses direction will build a very powerful analysis

    ReplyDelete
  11. Thanks all for your comments



    1- The data is shared already in "Data" section

    2- I added a section "FAQ" to describe the data values and the traffic density index

    3- Regarding accuracy, I think that the data and results are 80-90 % accurate, taking into consideration that bet2ollak website depends on users having smart phones with internet access which is only available to a specific segment of the population whom may have specific/different behaviors than others, but this was the only way for me to get relevant traffic data
    4- Finally due to other commitments, I wont be able to do more work on this research, but I am willing to provide the script used ( in addition to the data which is already published ) for anyone who would be interested in building on my work

    ReplyDelete
  12. Can the last Graph be made into an application or separate graphs so that I can study each road by itself ?
    Thank You :)

    ReplyDelete
  13. Sure, just send me your email and i will send you an xls file containing the data + overall chart so you can track each road separately

    ReplyDelete
  14. very good, thanks a lot

    ReplyDelete
  15. Mr. Karim, can you kindly make a recent study after 7azr

    ReplyDelete
    Replies
    1. I am sorry, i don't have enough time for that

      Delete
  16. This comment has been removed by the author.

    ReplyDelete
  17. this is the most useful analysis I've ever seen,many warm thanks for your efforts

    ReplyDelete
  18. Brilliant work......thanks and keep it up :D

    ReplyDelete
  19. This comment has been removed by the author.

    ReplyDelete
  20. اريد ان اقول للمهندس كريم عوده ... تسلم ايديك ... ويسلم مخك ... انت مثال لنوع جديد من شباب مصر ... خلقهم الله ليكون مستقبل مصر بين ايديكم ... ارجوك تمسك بالعمل الجاد والفكر المخلص لوجه الله ... لا تنظر للعطاء المادي فهو سيأتي باذن الله ... ولكن انت بدراستك القيمة لأزمة من اشد ازمات مصر وضعت عداد للحسنات يعمل ليل نهار ... يا بختك ربنا يحرسك ويرحم والديك اتمنى ان تراسلني على wafeeq@gmail.com
    ولك منى التحية والدعاء
    وفيق عزالدين

    ReplyDelete