Our achievements

Jungle Bus

Jungle Bus creates better data for transport mobility. The aim is to use this data to create the innovative services of tomorrow. Jungle bus collects and enriches data in collaboration with the Open Street Map community. The project is to set up the Daba’Go application in Abidjan based on data collected by Jungle Bus.

Daba’Go was chosen for its expertise in public transport especially informal in African countries. As part of a large-scale project in the Sahel region, the World Bank is working to map all public transport in the Mauritanian capital, Nouakchott, and to make this data public via open-source platforms such as OSM (OpenStreetMap) and Mapillary.

World Bank


Daba’Go’s innovative aspect, according to the OECD’s OSLO manual, is about product innovation. Indeed the application is a new good for the city and offers a new and improved service in relation to its characteristics and the uses for which it is intended. The solution responds to the real problems and needs expressed above in the context reminder.

If we also refer to the innovation classification table by Rebecca Henderson and Kim Clark, Daba’Go meets the criteria of modular innovation since it does not profoundly change the habits of transport users but makes important modifications on the tools they already know (Google Maps) to give another more adapted to the need.(Lopez, 2014)

In addition, Daba’Go offers a brand new algorithmic model for route calculation that is capable of combining traditional transport (tram, bus, train, etc.) and transport specific to large African cities (large taxi, small bus, boda, etc.) Using Artificial Intelligence/Machine Learning…

We go back “from scratch” on the route calculation systems to be able to consider transport that does not have a fixed departure and arrival schedule (large taxi, small bus, boda, etc.).

This is done in three stages:

1. Data Collection

Go to the field to retrieve the data associated with the transport in question.

2. A Deep Learning Approach

It goes perfectly with the problem of predicting the wait time of irregular transports which, at first sight, do not follow an apparent order but rather a hidden order that the algorithms of Deep Learning are able to find.

3. Incremental Learning

This method in the name of Incremental Learning and integrated into the application and allows the collection of data in real time ui come in turn to enrich and improve the algorithm.