Implementation of stochastic mobility services

In this Final Year Project (FYP), we explore the impact of stochastic variables on multi-vehicle mobility problems. Multi-vehicle mobility problems have been a prevalent area of research among the robotics and automotive research communities. However, many past literatures have either neglected or failed to address the presence of stochastic factors. Examples of stochastic fact are such as ad-hoc road works, probabilities of vehicle breakdowns, different performance of the vehicles and unexpected traffic lights. All those uncertain factors may prevent the vehicles from achieving the shortest travel time and adversely affect the overall performance of the mobility solutions. Hence, this FYP aims to understand the impact of stochastic variables and propose new solutions to tackle stochasticity. The FYP will be divided into three subprojects, where each subproject explores a different multi-vehicle mobility problem, specifically: (1) Multi-Vehicle Patrolling Problem, (2) Multi-Vehicle Locational Optimisation and (3) Traffic Light Aware Routing. For each mobility problem, we will study the solutions proposed by past literatures and discuss their limitations. We then propose a stochastic solution for each of the mobility problem, where we aim to optimise the solution while considering stochastic variables. Ultimately, we wish to determine a routing path that guarantees the least estimated travel time. Two simulators will be developed for subproject 1 and 2 to test our proposed solution and compare with other current state-of-art algorithms. The results from the experiments were encouraging. For subproject 3, we did a preliminary simulation based on real data from hands-on measurement. The simulation results had demonstrated the proof of concept of our approach and had shown that our proposed approach is superior to the current state-of-the-art algorithms.Bachelor of Engineering (Computer Engineering