IBM has selected my brain child (JOURNEY-GENIE) for Watson Build challenge 2018. We have been offered 7000USD IBM Cloud service credits in IBM cloud (Softlayer). This phase of the development is in PHASE 2 thereafter the solution will be pitched in San Francisco.
A marketplace solution for commuters using the Mass Bus Transit and Optimized Scheduling of Buses.
Lagos is the commercial capital of Nigeria, with a population of over 20 million people and Nigeria is the most populous black nation in Africa. There are rickety buses called “danfo” which the government has been trying to
eradicate.The newly procured commuter long buses are attached to respective routes within the metropolis.The commuters queue endlessly waiting for buses which arrive at the terminal after a natural cycle.
The solution would leverage on the patterns generated by the mobile application built for the commuters.
However, there will be data-integration of the tracking data of the buses with data generated by the commuters’ mobile application for planning/booking of the journey .
The commuters’ mobile application will be built using native Android and iOs in order to have excellent user experience and minimal dissipation of device battery.
The commuter will also have the opportunity to use a web application for booking/planning journey and also use “Journey Genie” which is a Watson assistant service built into the mobile application.
The commuter would plan the journey by booking at minimum of 30 minutes to the time before embarking on a journey. The data generated after the booking, would be filtered within the neighbourhood of 15 minutes prior departure time.
The machine learning algorithms (KMeans, KNearestneighbour, etc) will be applied to cluster, the population of commuter around the terminal/bus-stop with respect to destination terminal/busstop and Time of Journey.
Afterwards, the the commuter will be notified via a Push Notification service and SMS gateway service by implementing the Mobile First Development Kit to send the messages to commuters’ mobile application.
A configured client, implemented as backend adapter service to the Mobile First service in IBM Cloud will pull the output from the Machine Learning Algorithms saved into the db2 database service hosting the commuter journey itineraries and intermittently send the messages to the commuters using push notification or sms channel.
The algorithms will be improved over time, by evaluating the accuracy scores of the algorithm as the journey data increases over time.
Apart from planning the journeys, the commuter can also engage in the “Journey Genie” a menu within the commuters’ mobile application that make use of the Watson Assistant(conversation) to inquire for information on the arrival time of buses for specific departure terminal/bus-stop to destination terminal/bus-stop.
The mobile application will be integrated with payment engines, in order to pay for tickets. The payment reference will used to generate the ticket for each journey itenary. The receipt of the ticket booking will be shown from the mobile application before the commuter is allowed in to the bus.
The application will be built using the IBM Cloud. The application will be divided into four layers, namely:
1.Native Mobile application registered on the mobile foundation services in the in the IBM cloud.
2. The version of the mobile application will also be implemented in Web Application.
3. Suites of Adapters hosted in the mobilefoundation service in the IBM cloud will gather output data from the Watson Studio and the Watson assistance studio.
4.Web Layer/Application will be hosted on the IBM Cloud.
The web application wraps around the composite Watson assistant service and Machine Learning computation in the Watson Studio.
The Cognos service will be implemented to create dashboard on the output data from the machine learning computation.
The dashbourd will furnish the administrator/manager of the buses for optimal scheduling of the buses.
Commuters in Lagos, Nigeria using the new Lagos Commuter Buses
Poor scheduling of the Buses at rush hour
Poor ticket sales