Volcano
Descend into one of the world's most extreme environments with an immersive online volcanic expedition.
A World of Data
When it comes to the natural world, our scientific knowledge represents just a fraction of all there is to know. But in the age of big data, the opportunities to find insights that could lead to positive impact and a better understanding of the future are endless. By adding sensors and controls to physical objects, we can collect and expose more data sets – creating a smarter, more informed community.
The Power of Predictive Analytics
But data alone is not the answer. With the power of predictive analytics and machine learning, we can find patterns in just about anything, and build models to better predict future outcomes and trends.
Enter Predix, the Industrial Internet platform that lies at the intersection of physical power and digital potential. On Predix, we're able to connect big machines, data, and people to understand and identify issues before they even occur.
It is in this spirit that GE partnered with Qwake and the Nicaraguan Government to bring a volcano online and attempt to better predict its deadly activity. Together we launched a proof of concept endeavor to disrupt the way the world thinks about early warning systems.
Volcanoes
Experience the first step toward a more predictive system for the natural world.
1,571 Active Volcanoes
And while they’ve played a crucial role in shaping our planet and universe, the unpredictable nature of active volcanoes poses a potential threat to the estimated 800 million people living in their shadows.
What if...
Industrial Internet of Things (IIoT) technology could help inform communities about potentially dangerous volcanic activity before it’s too late?
Borrowing hardware and methodologies from other industries involving harsh environments, such as aviation and oil & gas, we set out to establish a new process for gathering, extracting insights from and sharing more accessible data – available to anyone and everyone interested.
Masaya
Known locally as the city of flowers, Masaya is a cultural hub and popular destination for tourists and explorers alike, who flock from near and far to experience the storied volcano from which the city gets its name.
With over 300,000 residents, Masaya is one of Nicaragua's most populous cities, but the potential impact of its namesake volcano goes well beyond the city limits.
Its close proximity to the nation’s capital, Managua, puts more than a third of Nicaragua’s ~6 million residents at risk.
Masaya is one of only seven volcanoes in the world with an active lava lake. Defined as large volumes of molten lava contained in a volcanic vent or crater, active lava lakes pose an unpredictable threat to those in surrounding areas.
"This one looked more like a waterfall than a lake. The government had already shut down access to the volcano."
Access to a lava lake this active provided a once in a lifetime scientific opportunity to capture data that has the potential to uncover some of our planet’s most inner workings.
Sensor Network
What does it actually take to connect an active volcano to the internet? It starts with a mesh network made up of 50 individual off-the-shelf sensors.
Sensors and their dataGathering all sorts of data, from atmospheric pressure and gas emissions to temperature and humidity, five different sensors built to withstand harsh industrial environments were sealed into ruggedized node enclosures and put to the test in some of the most extreme conditions on earth. Working closely with INETER, the team was also given access to seismic data via existing sensors that the government organization had previously installed and monitored.
A BIG PICTURE APPROACH
Measuring and analyzing various unique data sets together from a single node has the potential to tell a more complete story about complex natural processes like volcanic activity.
Ten separate sites
The nodes were placed at 10 sites within the volcano, from the crater rim, where base camp is located, all the way down into the lava lake, where temperatures regularly exceed 1000 degrees Fahrenheit.
A world class team of explorers, scientists and technologists came together to install the sensor network and carry out the mission.

The Drone Ninja
Radley Angelo

The Engineer
Hugo Nordell

The Astronaut
Scott Parazynski

The Catalyst
Sam Cossman

The Developer
Jayson DeLancey

The Scientist
Guillermo Caravantes

The Zipline Architect
Mike Thompson
The Mouth
of Hell
Using a true-to-scale digital model of the volcano created using drones, the team mapped out three geological levels to systematically plan their descent and the installation of each sensor node.
To get the crew and equipment safely into the depths of Masaya without damaging the crumbling walls along its interior volcanic crater, a multi-level zipline system was constructed – a true feat of engineering unto itself.

Ready to Descend
Cascading down over 1,200 feet, the world’s first and only zipline into an active volcano spanned from the outermost area all the way down to the lava lake at level 0.
THIS WAY TO LEVEL 2
Enter the crater
Level
Zero
This is by far the most dangerous part of the volcano, and where the truly valuable geological observations, lava sampling and data collection can be done. Two of the ten sensor nodes were installed here.
Level
One
This is a crucial zone because it’s close enough to the lava to detect volcanic activity, while being far enough away to avoid serious node deterioration. Three of the ten sensor nodes were installed here.
Level
Two
This is the entrance to the volcano and the only part tourists can access when conditions are safe. Five of the ten sensor nodes were installed here.
34°c
900m
Melting Point
Bringing technology into the Mouth of Hell had its challenges. Even so, most of the sensor nodes were able to survive for an average of 30 days before eroding due to the extreme conditions in the volcano. Yet the team was able to gather enough information to begin to see what the implications of these unique datasets could be.
Moving Toward a Deeper Understanding
By using previously untapped combinations of datasets, including historical data from INETER paired with real-time data and newly possible human observation, we hoped to get a glance into a world that was previously unknown. And we did. By analyzing the datasets on Predix, with its unique ability to harness vast amounts of unstructured data, the team started to see correlations between the data collected and their own observations in the Mouth of Hell.

THE SMOKING GUN
On August 16, 2016, the lateral vent of the Masaya Volcano opened, causing an increase in activity for the next 11 days.
STEP ONE TOWARD BETTER PREDICTIONS
During this time, patterns began to occur. When all was said and done, the team found a data anomaly at key sensors.
Explore how the different datasets were affected during this 11 day period.
AUG
16
AUG
27
uak
uak
uak
uak
uak
uak
uak
uak
uak
A Natural Anomaly
The simultaneous decrease in temperature and humidity at such key sensors is strange, as they typically have an inverse relationship. Seismic data and the team’s own observations, however, suggest this anomaly is in fact related to the activity in the Mouth of Hell.
Predictive
Analytics
In other words, based on the fact that gas emissions, temperature and humidity all went down at around the same time the vent closed, we can deduce that simultaneous increases of these measurements could potentially indicate future volcanic activity. And that’s where predictive analytics comes in. Working at the source of data, Predix can help detect patterns that go unnoticed to the human eye.
LOOKING AHEAD
From this proof of concept setup, we can start to see how sensor networks measuring dynamic datasets, paired with human observation and predictive analytics, could soon be used to develop a more predictive system.
Beginning
At GE, we don’t just think of incremental change. With Predix, we can integrate heavier, more varied datasets than ever before, and use them to make more informed decisions that help to push the world forward.
THINKING BEYOND MASAYA
We approached the Masaya expedition as a proof of concept program, hoping to inspire others to apply similar open data systems to more volcanoes in the future.
Thinking Beyond Volcanoes
And that’s still the hope, but it doesn’t end with volcanoes. As we collect more data about our earth and make it available to more people, we can explore other high-impact use cases for predictive analytics - from new types of early warning systems and beyond.
THINKING BEYOND MASAYA
When we work together, along with the power of predictive analytics, the possibilities for progress are endless.
Learn More
Interested in learning more and participating? Here are some ways you can join the community.
SPECIAL THANKS
A special thank you to President Daniel Ortega, Vice President Rosario Murillo, and the Nicaraguan Government for the vital support, forward looking vision, and scientific leadership, without which this ground-breaking technological expedition could not have happened. The institution responsible for volcanic monitoring in the country (INETER) also actively and decisively contributed through their scientists and resources, which under the direction of Dr. Wilfried Strauch, co-directed all scientific efforts and will be the main beneficiary of the technology left in Masaya for future scientific endeavor.