Small World connectivity in rural road networks

This study was based on GIS mapping of a taluk in Tamil Nadu. The mapping was done painstakingly of roads (cemented or mud) and village boundary (which is particularly hard to define, nobody really knows!) – most of this work is done visually.

The surprising math of cities and corporations. <TED talk>

The math of village structures and how it grows outs is equally fascinating and probably more diverse. Have a look at some satellite images of

Africa, very sparse

Africa, very sparse





India, numerous but small villages

READ MORE : Who is getting the jobs? | Kabaddi: An Honest Indian Sport


Ultimately to address questions such as

  • Will the addition of a new road dramatically change the livelihoods of the village residents?
  • How will a new idea/product spread through the rural network?
  • To map village road networks using existing maps to understand road connectivity in rural India.
  • Develop measures to estimate ‘urbaness’ of a region/taluk

Road connectivity information between 120 villages in Vadipatti Taluk, Tamil Nadu, INDIA. Road distance between nearby villages.


Analytics Method  
  • We had the distance measures between nearby villages. Using Djiktras algorithm to calculate distance between any 2 villages. Eg, Suppose we know the distance between places A & B and places B & C. Given this information, Djiktras algorithm will help calculate the distance between places A & C.
  • Dendrogram clustering to group villages by their nearness
Results Summary  
  • Designing Derived Variables : Distance between villages by road / As the crow flies distance might be a good measure of urbaness. To test this hypothesis we should measure this metric for more taluks (requires future work)
  • Surprisingly, we found a tight (somewhat linear) relationship. This means given the GPS location of 2 villages, we can estimate the road distance without actually having the road information.

Each color represents a village. This relationship is like to vary from terrain to terrain.

Each color represents a village.
This relationship is like to vary from terrain to terrain.

READ MORE : A Tale of Two Cities : Urban Transport in India | Locate, group and up goes performance

Connectivity of Villages
  • ‘Small-world’ kind of network with groups of 10-20 villages closely connected (nearby, blue) with each other and weakly connected (far away, red) to other villages.
  • A similar structure is seen in social networks captured by the famous theory of ‘6-degrees of separation’. Such a network means – travelling between any two villages should be quick (obviously assuming good roads!)
    Distance matrix estimated using Djiktras Algorithm

Leave a Reply

Your email address will not be published. Required fields are marked *