“Lives of Streeties” is a study of the street dogs in Bangalore, India conducted by me. Here I present a snapshot of the results of the study. The study is ongoing and will be updated regularly with new data. The term “streeties” is Bangaloreans fondly use to refer to street dogs.
The study is based on this premise:
- Street dogs (including partial street dogs* ) are free ranging city dogs.
- The behaviour of these free ranging city dogs might be the closest approximation of natural dog behaviour in cities (high density, busy environments)
- A large random sample of snapshots can expose patterns. These patterns shine a light on dog behaviour. This method is used because study of an individual dogs is logistically challenging and expensive due to the need of specialized equipment
* Partial street dogs are dogs that are being fed by a single home, given shelter in that home but are at liberty to walk in and out of the premises at any time and are not leashed or enclosed. They may sometimes sport a collar.
- Study street dogs to understand natural activity patterns of dogs
- Explore a viable way to study street dogs, which is:
- Low cost
- Easily scalable
- Share captured data for multiple analysis
- Present an avenue for “natural dog behavior” studies
- Dogs were observed and data was collected on the type of activity the dog was engaged in at any given point of time. Ex: Standing, trotting, foraging, walking, sleeping, sitting, lying down, barking etc
- The activity was then classified. Each activity was bucketed to one or more of these buckets.
- On feet
- Dog’s were identified where possible and named. This was not always possible.
- Charts created to capture patterns and trends
- Raw data uploaded for further analysis
Data Collection Methodology
- Data collected for each hour marker for 20 hours
- All data captured on video
- Multiple sightings of dogs considered multiple data points
- Day and time recorded for data point. Temperature, weather conditions and traffic conditions recorded for each data point
- Change of activity as a result of the data collection has been omitted from the study
- More than 400 data points (videos) gathered
- Data gathered on 14 different days
- Data gathered across 13 different temperature points (21C – 32C)
- Data gathered in different weathers – sunny, overcast, light drizzle, steady rain, after intense rains
- Study done on 38 uniquely identifiable dogs and several more unidentified dogs
All the data collected was broken down into different buckets. Three high level questions were asked:
- How many of the total observed dogs were awake and how many were asleep?
- How many of all the dogs observed were actually on their feet?
- How many of all the dogs observed were moving on their feet?
The results are presented here
Awake Vs. Asleep
Note that of all the dogs observed 40% were engaged in one single activity – sleeping. The remaining 60% of the dogs were engaged in all other activities like standing, sitting, trotting, foraging, pooping, barking, begging etc. The biggest chunk went to a single activity – sleeping.
Dogs on their feet
When this question was asked the graph flipped around. Only 40% were on their feet. Of all the dogs observed almost 60% were not even on their feet. They were either sitting, lying down or curled-up and sleeping.
When the question asked was “How many of the total dogs observed were moving”, the results were quite stark. Only close to a quarter of the total dogs observed were moving. The moving included a host of activities like walking, trotting, sniffing, foraging etc.
For those interested in getting a different view of the activities a dog is involved in, the activity cloud is a good visual representation.
Note that the most prominent activities are sleeping, curled-up and resting. The next set are standing and walking. Then comes trotting and sniffing. All other activities are far less frequent in comparison.
Since the data was collected throughout the day, the data was charted across the day to see if there were discernible patterns across the day. The data has not been charted across temperature, traffic conditions or human movement. While this graph might be influenced by all of these factors, at this point the graph only exposes how the activities look across the day.
Awake Vs Asleep
Notice that somewhere around 8:00PM the graph significantly changes, with the number of dogs being awake being far more than the number of dogs asleep.
Dogs on their feet
In this graph, the gray line represents the number of dogs asleep. Note that in the begining of the graph the gray line is in the lower half and the blue line is on top. But around the midnight mark this changes and the gray line starts spiking up indicating more dogs on their feet after this point. The 2:00 AM data is a outlier because of heavy rains and needs more data.
This graph is just a more clear indicator of the same phenomenon that the previous graph exposes. The number of dogs moving is quite low in the first part of the graph. Then it starts spiking somewhere around the midnight mark.
- The most common activity that dogs are involved in is sleeping.
- Of the awake dogs, there is not much heavy activity. The activity seems light. Very few dogs are moving. The moving dogs are include a host of activities, none of them are very physically intensive activities.
- Dogs seem get more active after dark.
- Data collection without influencing the subject is very difficult
- Night time data gathering is a challenge in terms of safety and equipment
- Unique identification of dogs is challenging
- Plenty of data required to get clearer patterns.
- Weather variation not enough to observe weather related patterns.
- Unseasonal rains interfered with data collection.
All Raw Data
This youtube playlist contains all the videos captured as part of this study and will be updated with more videos as captured. Feel free to use it for further analysis. Please inform me of any uses of these videos, either for analysis or other purposes to help mould the future of this project to make it more useful to dog behaviour enthusiasts.
- Gather more data points to iron out discrepancies
- Identify other uses for the data gathered. Inputs are welcome on how the data can be used.
- Identify other kinds of data that can be gathered in this fashion
- Get funding for scaling up project:
- Data gathering can be scaled up tremendously if people could be paid to gather more date
- Expand study to rural (if there are behaviour differences)
- Obtain specialized equipment for night study