 Our first speaker, Milad Namarzadeh, who will be talking to the title probabilistic learning and planning framework for optimal management of systems under uncertain environments. Welcome Milad. Every day in our life we use infrastructure systems whether we're aware of it or not. Systems like roads, highways, water and wastewater pipelines, wind energy, power grids and the failure of these systems or the lack of service to the damage to them could be really catastrophic or costly. For example, in 2009 a wind turbine collapsed in Fenner Farm in New York State. According to our industry collaborator, this failure was costly and it was due to the fatigue in the mass foundation. So how can we stop it? There's a need for careful management of their operation and maintenance and here is where my research comes in. We specifically posed three research questions that we have tried to address in my research. The first research question involves the link from the information node to the knowledge node. Given the information collected from the infrastructure components, how can we learn their degradation behavior and also the effectiveness of the maintenance actions? For example, in this setting we use a probabilistic method based on Markov chain Monte Carlo to be able to learn the behavior of components. Furthermore, to extend this to the network level where observation of one component is not only relevant for that specific component but also to learn the degradation behavior of all components on the network. We use the hierarchical Bayesian modeling approach to allow the flow of information among the components on the network. The second research question corresponds to the link from the knowledge node to the decision node. As a manager operating these infrastructure systems, you want to know what is the best thing to do at each time a step considering the huge uncertainty caused by the environment. So we propose the method called PLUS which stands for planning and learning for uncertain dynamic systems which tries to identify the optimal action given the uncertainty that is caused by the environment. The third research question corresponds to the link from the decision node to the information node. If you are managing a network of 100 components and you can only collect data from 20 of them at each time, how can you prioritize this task? To be able to do this, we quantify the value of each information collected from the components using a pre-costerior analysis to be able to prioritize the data collection approach. So overall, we have proposed the framework for optimal management of infrastructure systems that integrates learning, planning and data collection scheduling. Thank you very much.