Network Analysis of Matatu Routes

Yesterday’s Data Science meet-up at One Africa Media had an interesting presentation from Henk Harmsen. He picked the data set from digitalmatatus.com and built a network graph of bus and matatu stages in Nairobi.  In it, he uncovered fascinating facts. First, which is the busiest matatu stage in Nairobi?  It’s not what you think – turns out Agip the nondescript stage next to KRA handles about 80% of matatu traffic in Nairobi. The diagram below captures the details.

Nairobi Backbone Transport Network

Nairobi Backbone Transport Network

Seocond. While Agip handles most traffic, Ambassadeur stage provides the best interconnection within the city i.e Ambassadeur is  the best axis point of connecting a traveler from one part of the city to another.  It is worth noting that Ambassadeur stage host the three largest bus fleet in Nairobi  (KBS, Citi Hoppa and City Shuttle) as per my previous analysis. A further confirmation that these bus companies have the widest coverage of the city. Let’s look at a different perspective – if Agip and Ambassadeur are removed from the list of stages, which are the next busiest stages in Nairobi? The diagram below points to Khoja and Kariakor.

Route Network in Nairobi

Route Network in Nairobi

These visualizations poke a hole in the notion that most traffic in Nairobi CBD emanates from Eastlands. Seems most traffic in the CBD originates from Mombasa road, Ngong road and Thika road respectively.

Visualizations by Henk Harmsen

 

Advertisements
  1. An interesting insight but unfortunately it misses the reality on the ground such as the fact that mapping locations within the CBD is misleading, for example the Agip stage serves matatus from 3 different directions namely Langata Road, Mombasa Road, Ngong Road and Jogoo Road.

    In addition the Agip bus stop only handles inbound traffic its opposite stop handles almost nil traffic thus if I was new to Nairobi and followed your analysis I would spend the entire day at Agip waiting for an outbound matatu.

    I also disagree with the assertion that Ambassador has the highest connections, please visit the central bus station (Bus Station) and you will realise that you are wrong.

    Blindly analysing data can result in serious misinformation as has been clearly seen from this article, without feet in the ground data is useless in generating useful information

    Like

    Reply

    1. Hi Robert,

      I think you’ve misinterpreted the analysis. The first issue on Agip, the fact that it handles traffic from 3 major roads already indicates that it processes more matatus than any other stage.

      On Ambassadeur, I’ll also say you are wrong. The meaning of the analysis is that from Ambassadeur stage you can move to different parts of the city e.g Kawangware, Embakasi, Karen, Buru Buru e.t.c while Bus Station has less options to move to different parts of the city.

      Like

      Reply

  2. Hi,

    Which 3 major roads are you referring to as relates with Agip?

    Central bus station handles traffic from Jogoo Road, Thika Road, Ngong Road, Mombasa Road and Langata Road.

    This article on Amazon might bring home better the reason why extracting information while sitting in the cloud can be detrimental to your business.

    http://www.bloomberg.com/graphics/2016-amazon-same-day/

    Like

    Reply

    1. Hi Robert,

      One thing you forgot it that all matatus/buses that go to Central Bus Station from Mombasa Road, Langata Road and Ngong Road have to pass by Agip. In addition, the data utilized in this analysis was not gathered from the “cloud” rather by on-the-ground research, where field assistants boarded all matatu routes in Nairobi while recording the geo-location. You can read about it here:
      View story at Medium.com

      Like

      Reply

  3. Killion Mokaya May 22, 2016 at 7:10 pm

    The classification of Agip as a matatu stage is flawed. It is more of a dropping point. Realistically you will not board a matatu there to the city’s outskirts and matatus do not spend and average of more than one minute at Agip. Consequently the findings are misleading.

    Like

    Reply

    1. A drop point qualifies as a stage – perhaps you meant a terminus in which Agip isn’t. The findings aren’t misleading rather understood differently.

      Like

      Reply

  4. […] analysis results can be found on this Portal . There has been vigorous efforts by the government to ease traffic within our city but has not yet […]

    Like

    Reply

  5. Great work. Suggestions though is that the actual frequency of matatu traffic per stage would present a clearer picture of how busy each stage. This can be clustered into different colours for easier interpretation.

    Like

    Reply

    1. Thank you! Will see if I can redo the visualization.

      Like

      Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Advertisements
Advertisements
%d bloggers like this: