WEBINAR
Predictive Modeling of Dredging Needs
Each year over $100 million is spent to dredge southwest pass and due to current methods, these dredges are not proactively dredging. Our goal is to increase the ability to dredge before the shoaling accumulation impacts shipping and transport needs of the ports in the Mississippi River.
By increasing the window of prediction from what is currently 14-days to upwards of 45 days we could help reduce the need for emergency dredging and also prevent downtime in river traffic. Our model has evolved from a 1-dimension volumetric model to a 2-d approach pinpointing the regions within the river that need to be dredged.
We plan on transitioning from a recurrent neural network to a convolutional approach using historical depth data of multiple reaches within southwest pass.
The presentation will include discussion of the techniques used to process the data, the data collection and the ideas present in building the models.
September 18, 2024 | 12-1 PM EDT
Attend through Webex by clicking the Register Now link above. Please check your spam folder if you do not receive an email confirmation from Webex after registering.
Speaker – Hans Jacob Broders
Hans Jacob Broders has a background in physics and mechanical engineering with a focus on physics-based simulations. With ARA for the last 3 years, he has transitioned into a role as a data scientist focusing on building predictive models to help understand real world phenomena. Mr. Broders recently completed his Master of Applied Data Science from Syracuse University and has a strong understanding of the tools relevant to creating informative predictions and presenting strong visuals to help with understanding of the information.
