During my time at 4C, I have worked on a number of supply chain optimisation initiatives. Each of these has required substantial data analysis. To give you an example, our latest project required the analysis of a quarter of a million point to point movements. The growing availability to this type of data has already allowed 4C to drive significant benefits to a wide range of clients. This is, however, just the tip of the iceberg.
As we gain access to more and more data, we will be able to deliver increasingly optimised solutions across a range of categories. In this context IBM’s recent paper, which was part of Orange “Data for Development“, offers a small glimpse of the future.
Getting Everyone Aboard
Orange “Data for Development” is an open data challenge, encouraging research teams to use datasets of anonymous call patterns of Orange’s Ivory Coast subsidiary, to help address society development questions in innovative ways. The project provided research teams with an incredibly rich dataset of 2.5 billion anonymous calls and SMSes sent in the Ivory Coast over a period of five months.
IBM’s research team first isolated data for the country’s capital, Abidjan. The city has seen a tremendous population growth over the past decades and public transport infrastructure has not been accordingly optimised. The challenge the team set themselves was to develop a system, called “AllAboard”, to analyse the patterns of user movement and identify likely home and work locations for individual users.
Optimising Abidjan’s Bus Network
As the data captured the approximate location of a large number of users, IBM was able to extend its analysis beyond the current use of public transport and understand the true start and end point of each journey. This analysis allowed the system to propose two scenarios.
The first consisted in optimising existing routes to match the frequent movements made by users. The second proposed additional routes that could be added to build the customer base of the public bus operator. These additional routes are currently undertaken by smaller vehicles which cause significant congestion and carbon emissions, as well as being poorly regulated.
The IBM team concluded that if the data could be combined with information about individual users, it would be possible to create routes targeted at those most likely to rely on public transport (i.e. by excluding car owners). The analysis identified an overall opportunity to reduce average travel time by 10 per cent.
Of course this data has a number of limitations – users are selected due to phone usage and network, and locations are approximate as they are based on the nearest mobile phone tower. However, more sophisticated data sets already exist. Google’s Location History, for example, tracks devices using GPS and provides far more accurate information.
Although the legal implications of using this type data have yet to be fully understood, what is clear is that this information will lead to significant opportunities. From designing more efficient transport networks to implementing optimised commercial strategies, businesses have only begun to scratch the surface of the possibilities on offer.