Abstract
Urban traffic congestion is a huge problem that affects cities worldwide, particularly in rapidly
urbanizing areas where population growth significantly outpaces infrastructure development.
Kinshasa, the capital city of the Democratic Republic of Congo (DRC), exemplifies these challenges,
with inadequate road infrastructure unable to meet the needs of its growing population.
This has led to severe traffic congestion, increased pollution, and substantial economic losses,
necessitating the need for effective solutions.
Presented in this thesis is a novel approach to managing traffic congestion in Kinshasa
through the integration of multiple methodologies — COMPRAM (Complex Problem Handling
Methodology), System Dynamics, and Artificial Intelligence (AI). COMPRAM is used to comprehensively
address the multifaceted and interrelated nature of traffic congestion, involving
phases such as problem identification, stakeholder engagement, scenario building, and solution
development. System Dynamics modeling is employed to understand the feedback loops, delays,
and causal relationships within Kinshasa’s traffic system, offering a dynamic perspective of how
different interventions can impact the city over time. AI techniques, such as Large Language
Models (LLMs), Physics-Informed Neural Networks (PINNs), and Extreme Learning Machines
(ELMs) are also incorporated to assist either in the planning process or in developing adaptive
traffic management solutions capable of predicting and mitigating congestion in real-time.
The study identified key factors contributing to traffic congestion in Kinshasa, such as rapid
population growth, deteriorating infrastructure, and inadequate public transportation. The
research models and of the various traffic scenarios by employing System Dynamics, provided
insights into how different interventions — such as increasing public transport capacity or
improving road infrastructure — could alleviate congestion. The employment of AI-driven tools
showed that the potential to further enhance these models by providing predictive capabilities,
allowing urban planners to make informed, and data-driven decisions.
This interdisciplinary approach aims to provide a sustainable and cost-effective solution for
Kinshasa’s traffic issues, particularly in light of the city’s limited resources. The findings of this
research demonstrate that a combination of structured problem-solving such as COMPRAM,
dynamic system modeling, and AI can significantly enhance the effectiveness of urban traffic
management. Moreover, the proposed methodology offers a scalable framework that can be
adapted to other cities facing similar challenges, providing a pathway toward more intelligent,
adaptive, and resilient urban transportation systems.