Tech for Good

For example, when the killing of Africa’s savanna elephants accelerated dramatically due to increased global demand for ivory, Paul funded an unprecedented aerial survey aptly named the Great Elephant Census. It was the first of its kind in more than 40 years, and the first coordinated, standardized count ever attempted. It brought together more than 90 scientists, 286 flight crew members, and various NGOs to survey nearly 600,000 square miles in 18 different countries. 

The project’s success, and its sobering results, led his team to create and deploy EarthRanger in 2016. It’s an online software solution that lets park rangers track movements of wildlife, rangers, vehicles, and intrusions in real time — no small task since the regions being monitored are vast (in many cases, tens of thousands of square kilometers). But with EarthRanger, the data from hundreds of digital radios, animal collars, vehicle trackers, and sensors are compiled and transformed into actionable graphics and visualizations. This gives officials and protected area managers the ability to deploy resources quickly and efficiently when they’re needed and where they’re needed. For example, if a snare is discovered or an elephant enters a human settlement, the closest rangers can be dispatched immediately.  
 

Since it was first deployed at Kenya’s Lewa Wildlife Conservancy in 2017, EarthRanger has expanded dramatically. It now monitors wildlife reserves in hundreds of sites across six continents as it tracks the locations of over 50 endangered and vulnerable species.  

And, as the community of users continues to grow, so do use cases. This new technology is now helping to protect endangered white rhinos in the Serengeti, watching for illegal deforestation in Southeast Asia and combating locust swarms throughout East Africa. This steady expansion in vastly different locations (from the North American prairie to the coast of Mozambique) shows what a powerful enforcement tool EarthRanger is for protected area managers everywhere. 
 

Another similar project is Skylight. Paul and his team created it to address the huge problem of illegal fishing that runs rampant in ocean waters around the globe. Because of illegal fishing, billions of dollars are lost, human rights violations go undetected, and millions of lives are impacted. In 2020 alone, 93 percent of the world’s major marine fish stocks were either fully exploited, overexploited or significantly depleted because of unsustainable, unreported, and/or illegal fishing practices. 

With Skylight, maritime analysts are empowered to identify suspicious vessel behavior and alert authorities who can take enforcement action. By watching for suspicious behaviors like transshipment (transferring cargo from one vessel to another while in transit), refueling, and entry into protected areas, Skylight makes it possible to dispatch patrols within minutes. 
 

Skylight also compiles millions of incoming data points, including those from state-of-the-art radar systems. It then analyzes them using advanced machine learning and checks all findings against a global vessel registry database. 

The system currently monitors 16 million square kilometers of ocean in 19 countries. Recently, it was instrumental in a successful, joint-operations venture in Ghana where the Ghanian navy, Skylight, the UN Office of Drug and Crime, and Global Fishing Watch collaborated to investigate 14 suspicious vessels. The action resulted in the arrest of individuals aboard four ships who were engaging in Illegal, Unreported and Unregulated (IUU) fishing.  
 

Closer to his home, to protect the last remaining 75 Southern Resident killer whales in Pacific Northwest waters, Paul’s work advancing machine learning and artificial intelligence (AI) laid the groundwork for the 2021 Aquatic Mammal Photogrammetry Tool (AMPT).  

In the years since his passing, his team at Vulcan has worked with Dr. Holly Fearnbach and Dr. John Durban who have been using aerial photogrammetry for more than a decade to assess the health of endangered Southern Resident killer whales.  

Before AMPT, images had to be processed manually, by humans, using datasets — which takes anywhere from four to six months to complete. But by using machine learning and an end-user tool, the time needed to process images dramatically decreased to a matter of weeks, or even days. This faster turnaround of health metric data is helping guide quicker management decisions like targeting specific salmon runs or limiting disturbance by vessels with the goal of improving accessibility to Chinook salmon, the whales’ primary source of food. 
 

Paul’s use of technology for good has also had an impact on work being done to address climate change. His legacy can be seen in Vulcan’s partnership with NOAA’s Geophysical Fluid Dynamics Laboratory (GFDL). The team is designing climate models to run on the world’s largest supercomputers in high resolution. These technologies are dramatically increasing climate model speed and accuracy by replacing older components 

By providing valuable insights and predictions to help people understand what changes are on the horizon, populations everywhere are better able to prepare, and policy makers are better equipped to make decisions and address challenges for the benefit of people the world over. 

EarthRanger, Skylight, and both the climate change and machine learning teams all continue their global impact to this day, now as part of the Allen Institute for Artificial Intelligence (AI2) in Seattle.