The frigid waters of the dark and briny deep may first appear uncharted, but in reality the data is out there, scattered across public sources. Getting three dimensional data from a few leagues of the sea can be a nightmare for researchers and industry professionals alike. The 3 petabytes of public ocean data out there is disconnected, often archived and never used again. Making even a fraction of that more accessible offers a huge value to anyone looking for it.
On top of that, once you find the data it is still difficult to work with. Mariexplore, a Tallinn/San Francisco based company estimates now that 80% of exploration time is spent on data processing. Marinexplore wants to cut that number fivefold. To do so, they have brought together a huge chunk of the publicly available data and release it into an easily tappable resource, while simultaneously building a community around their data.
“The problem with all of this is that the ocean and seas are very big, and no one can own all of the hardware in the world to gather it,” Marinexplore founder and CEO Rainer Sternfeld explains.
At launch, Marinexplore offers more than 350 million measurements from 23 000 devices. These data are provided by public sources like NOAA stations, GTS buoys, oil platforms and drifters, JCOMMOPS Argo Program, Rutgers University gliders, Liquid Robotics PacX Challenge data, and others.
“The first problem we have solved is that we have created a tool that enables to search, find, and export ocean data faster and easier than any other tool out there,” says Sternfeld. “The main thing is that we want to become the number one footprint on the web for oceanographic data, bringing together all the public data.”
In use, Marinexplore has put together a very simple interface built on top of Google Maps, where there they offer a three dimensional data structure where the data is positioned in geo-space. Scientists and industry professionals can easily grab chunks of three dimensional data without having to find and process the data themselves.
Accessing the data is free, and in the beginning Sternfeld predicts that their customers will be oceanographers to reseachers, to geo-physicists, marine biologists, and many others in the public and private domain. On the private side, he predicts shipping, insurance, and stuff like that will also find their service useful.
To monetize that the company is looking into building data streaming products. They are building an API where you would be able to define an area in the seas and then only get that specific perimeter from that specific area. Potential customers include weather forecasting systems, shipping companies, and others who would like to predict better current models, risk models, and so forth.
“Data itself doesn’t mean anything, right?” Sternfeld says. “Our intention is to build on top of machine learning systems so that we can actually provide forecasting products eventually. But for right now, we are providing very simple access to that data in an aggregated way.”
The company has moved part of its operations to Silicon Valley, which initially seemed strange because the startup seems like it doesn’t need to plug into the same ecosystem as your average social app. But Sternfeld tells me there is actually a lot of oceanographic talent to draw from in the Valley, and one way or another they needed to be on the east or west coast where the ocean community lives. Marinexplore is building a technical team of ocean modeling engineers, 3d visualization guys, and so forth. In the Valley, the company then has access to talent from the Jet Propulsion Lab, as well as the Google Earth guys that have worked with these challenges before.
The startup is definitely interesting for its societal value to the oceanographic community, and it will be interesting to see how Marinexplore will be able to expand as a central hub for ocean data, as well as connect knowledge between industry professionals and scientists. When talking to Sternfeld, he called themselves almost a git-hub for ocean data, which appears to be an apt metaphor of what they’re trying to do. How many other industries could benefit from a central resource and community around data?