The purpose of this graduation thesis entitled "Probabilistic restructuring of complex construction project activity linking using GERT: an alternate network distribution in mitigation controller" is to analyse the effect of project network structures when the duration of the project activities is distributed stochastically and its following effect on the existing Mitigation Controller©. The researchers at TU Delft have developed a state-of-art tool called Mitigation Controller© (MitC) for automating the search for finding the most cost-effective set of mitigation measures to ensure the probability of the project's completion at a required level. However, there is a modelling error in the project network diagram of this tool. The risks and uncertainties are modelled using a Program Evaluation and Review Technique (PERT) distribution. The current Mitigation Controller has managed
to recreate the human-oriented selection of mitigation measures, but it has not considered all possible scenarios within a complex construction project.

PERT is the most straightforward distribution used for explanatory purposes. This does not reflect reallife conditions. There are instances where certain activities are repeated when the result does not meet the required quality. To overcome the limitations of PERT, it is suggested to implement the Graphical Evaluation and Review Technique (GERT) for scheduling construction projects. A new code
is integrated into the existing Mitigation Controller to generate new network paths based on the probabilistic nature of the project activity. The novel network structure is used to compute the optimal set of mitigation measures using Monte Carlo analysis and linear optimisation. It is also observed that the optimisation solver takes a substantial amount of time to compute depending on the number of activities on the project. An efficient Monte Carlo analysis is implemented to reduce optimisation time within the tool.