- Stakeholder: Trackmatic (partner of Famous Brands)
- Students: Mponeng Ngwenya, Jurgens de Bruin, David Petrus Van Niekerk (UP), Khehla Malakoene (NWU)
- Project Lead: David Slotow, Dewald Lindeque, Quentin Williams
- Project Mentors: Nkosinathi Ndlovu, Albert Dove, Ndamulelo Netshiavha
- Year: 2016/2017
The Famous Brands is a project initiated by Trackmatic a partner of Famous Brands that tracks the fleet of Famous Brands during distribution. Trackmatic business collects a lot of data that when analysed and visualised correctly, can assist their partner to make informed decisions to improve and grow the business. The objective of the project is to assist Famous Brands to reduce costs, maximise resource utilisation and improve performance by pointing out inefficiencies hidden in the data. Famous Brands team developed a framework that measures each site’s Key Performance Indicator (KPI) using various route variables, then drilling down various factors including driver behaviour ( such as late deliveries, missed stops etc.), planned vs actual information and trucks (including but not limited to duration and distance).
The team has used the KPI to compare the overall site performance, and the ability to drill further will indicate what the business needs to concentrate on. To ensure ease of implementation a multi view approach was used to visualize the data which consisted of a sunburst showing KPI’s, truck and the driver graphs as drill down per site. A map visualization was created in which the geocoordinates of Famous Brands Sites were plotted for site selection. A selection of a site on the map will show the KPI measure for the site. From the data, they also wanted to portray truck utilisation by day of week, highlighting the use of hired trucks per route. This could be a powerful representation for future planning. Overall, the Famous brand team endeavoured to highlight discrepancies that could be used by decision makers.
The students felt that the project gave each of them the opportunity to learn things that they were not exposed to considering there was a lot of data analysis that was required. They learned that lack of data meant thinking of ways to fill up their data. The Project- Famous Brands project team used google to compare the routes given to them and the best that could have been used instead. They hoped to use machine learning for their project however there was no way of doing that with the kind of database they had. The DSIDE program exposed them to several Data Science tools and libraries, these range from data visualisation libraries like C3 and Highcharts, incorporating openlayers for an interactive map visualisation. With regard to machine learning scikit-team Was used as the primarily tool, Project-Famous Brands team felt that their thought process would be the most value in future projects.
Author: Team + Nolihle Gulwa, B Tech Journalism, Walter Sisulu University.