The principal scientist at Carnegie Mellon University, Christoph Mertz, is onto understanding landslides – prediction and prevention. Dealing with decoding the natural phenomenon, the project centers around identifying signs of forthcoming landslides. Once, fully developed, it would help the American Economy save big bucks spent on rehabilitation. Besides, it would help prevent the loss of lives in the communities at the brunt of it.
Christoph Mertz – Machine Learning to Pave Way:
As part of his endeavor to improve the current situation, he takes pictures of the hills overlooking Pittsburg’s West End. Mertz, in partnership with Alleghany County, is currently examining five potential sites of the phenomenon to understand system viability. The idea is to identify patterns that can help put signs of an impending landslide into perspective. Pictures will help in the task by highlighting changes. Few signs that give away include dents in road fences, the inclination of foliage, and debris on road.
According to Mertz, for the project to succeed it is imperative to build historical data and even comprehend geology pretty well. Deep learning algorithms function in the data fed to them. It is quite similar to how Google uses linguistic data to translate and differentiate French from Spanish. Therefore, a strong database is imperative to building a viable machine learning system in the future. This, in turn, calls for an inter-disciplinary approach to the issue.
What the Project Brings to the Table
A win-win situation will be created once the project is fully developed and brought to life. With the data collected and analyzed with the aid of machine learning, the prediction would be possible. Till a few years ago landslide prediction wasn’t even a part of the vernacular and this project makes it a possibility. This will help the government act swiftly to help prevention, saving both lives and money. To add on to it, the project aims at understanding and explaining factors that would shape policy and budget allocation. Also, often there is a budgetary bias that leads to a de-prioritization of areas with marginalized populations from resource allocation. The goal of the task is to provide authorities with objective information so fairer allocation can be made.
