Location Intelligence | Pitney Bowes

Using Isochrones and Isodistances in your Big Data Analysis

By Rose Winterton

Isochrones and Isodistances have long been the mainstay of site selection in producing store catchment areas. Potential store locations can be assessed against travel times for potential customers, then overlaid with geodemographic data, competitor store locations and other variables to produce a site selection model. Retailers can optimize a location based on variables that drove profitability in their existing store network.

However, the uses of routing data and routing algorithms go beyond site selection, and in the era of big data new uses cases are emerging. The power of big data comes in processing data fast enough that answers to complex problems can be pre-populated rather than assessed on the fly. Pitney Bowes uses these techniques to process data and produce value added products. One great example of this is our fire protection risk data bundle. Our routing engine is used within a big data framework in order to calculate the three nearest fire stations to every property in the United States. Running the analysis in Pitney Bowes environments means that customers can benefit from the power of big data without needing to understand big data processing. Pitney Bowes provides the output results which the customer can then use as a simple look up table.

A recent use case involves planning for a future network of electric cars. Today one of the limitations of electric cars is the battery capacity and range leading to the need to recharge at a limited set of charging stations. As countries such as France, Germany, UK, Norway and India look to drastically reduce the number of petrol and diesel vehicles in favor of electric, these challenges will need to be overcome. Part of the answer lies in building out a network of charging stations. Information on typical journey start and end points coupled with population data and drive times become datasets which can be used to generate a site selection model. Processing these datasets within a big data framework allows quick feedback to the model with data scientists being able to iterate and evolve the models with a rapid turn-around.

Check out our latest Spectrum for Big Data release including isochrone calculations.