David Kerkkamp wins ICAART Best Student Paper Award
The name of the winning paper: “Grouping of Maintenance Actions with Deep Reinforcement Learning and Graph Convolutional Networks”
At the 14th International Conference on Agents and Artificial Intelligence (ICAART), David Kerkkamp was awarded with the Best Student Paper Award. David’s work was supervised by TU/e researcher Zaharah Buksh, who, together with Yingqian Zhang, are the co-authors of the winning paper.
The aim of the work was to solve the problem of optimal maintenance planning using historical data. Instead of dispersed maintenance actions of individual pipes across the underground sewer pipe network, the work proposes an approach that combines Deep Q-Networks (DQN) with Graph Convolutional Networks to ensure the grouping of maintenance actions of geographically close pipes