Royal Mail is the UKs national postal service, delivering approx. 15 billion items to nearly 31 million addresses annually. Despite handling more than 50% of British parcels, Royal Mail is on the look out for efficiency in this extremely dynamic and competitive market. While protecting operational cost and enhancing customer satisfaction are our main drivers, designing a new dedicated parcel delivery service from scratch is no small feat…but the Data Science team is here to help!

Route optimisation technology has now become invaluable in the logistics market and is driving high impact in the Last Mile delivery journey. Based on a long academic history that started with the well known Traveling Salesman Problem, route optimisation is one of these NP-hard computational problems that requires creative algorithms to extract value from an extremely challenging combinatorial problem. While it is true that there is a range of very good software providers in this space, there are also many incentives to develop an in-house capability, most notably the ability to build a tailored solution for the specific postal constraints.

Building on last year presentation of the general problem, this 3 hours interactive session will allow us to go through the entire cycle of a PoC project, covering stakeholder requirements, data acquisition, algorithmic modelling as well as service delivery review. In particular, we will show how heuristics can be crafted manually to match up with a standard VRP library.

Target Audience:

This workshop is designed for all kind of Data Scientists, from hand-on developers to more managerial roles.


• Basic python development environment (pip, python=3.6)
• Standard libraries, such as pandas and numpy
• We will also make use of Google OR-Tools library:
• Finally, highly beneficial if attendants are familiar with last year’s DSF talk from Royal Mail. Record can be found here:

Supported by