Innocent Ndibatya

Modelling Organically Evolved Public Transport Systems to improve efficiency: An Intelligent Complex Adaptive Systems Approach

PhD Student

Public transport systems in developing cities of Africa, Asia and Latin America are inefficient and dangerous to the travellers. These Organically evolved Public Transport Systems (OePTS) are synonymous with  unprecedented delays, dangerous traffic accidents and heavy pollution, yet they transport majority of the working population. This has led to; (1) low productivity due to time lost, stress and accidents during travel; (2) increased carbon emissions due to high miles per vehicle per trip, high engine run-time, and advanced age of the vehicle fleet. Consequently, inefficient public transport systems have negatively affected the economies of developing countries.

The main goal of this PhD research is to improve efficiency of public transport systems in developing cities. This will be achieved by using intelligent transport systems (ITS) technologies and artificial intelligence (AI)
methods to support agents/actors in the OePTS to make better optimised and intelligent local travel decisions at a micro level. The hypothetical basis of this research is; agents/actors in OePTS are governed by simple rules at micro level. They synchronise their behaviour through a process of self-organisation that emerges from the micro to macro level. By using ITS and AI to support the agents’ decisions and behaviour synchronisation at micro level, a better optimised state emerges at macro level. The new emergent state is more efficient to the travellers, transport providers, and is environmentally friendly.

During this PhD research, Intelligent Complex Adaptive Systems approach will be used. This approach will involve studying the dynamics of public transport in Kampala by agent-based modelling (ABM) and Machine
Learning to discover the rules that govern OePTS agents at micro level. The agents’ decisions at micro level will then be synchronised and better optimized decisions learned through an adaptation process. The new emerging state of the system will be evaluated to ascertain if efficiency has been achieved.