Modelling Urban Traffic Data into SensorThings

As cities grow, traffic congestion becomes impossible to escape. Commuters become impatient while stuck in traffic wondering where the accident has occurred, how far you are from it, and is there an alternate route. Expanding roads or making alternate routes are billion dollar projects that will span over years. A more cost and time effective solution would be to keep the users informed in realtime about traffic accidents and road closures.The city of Calgary has created a realtime open data feed for traffic accidents and road closures. Lets take a look on how we can model traffic data into SensorThings.

Figure 1 shows visually at a high-level how we extract and parse the data received from the sensor feed in order to feed it into SensorThings. Before we push the data into SensorThings we have to create the respective entities. Figure 1 shows in detail how the data is modelled into SensorThings using traffic accident example.

The data can be modelled in SensorThings where a Thing represents the road which will have a static location. Since the location of the road is static we will only have one location and one historical location in this example. A Thing that can have multiple Datastreams like road closures, traffic accidents, etc. Each Datastream will have one sensor and observed property. The observed property like the name depicts, is what property you are observing in that Datastream. The sensor is an instrument that observes this property. From there we can push our data as observations into the respective Datastream. SensorThings can model any type of data received from a sensor. It also allows developers to easily build and deploy IoT applications.