Spotting Objects From Space Is Easy. This Challenge Is Harder

Evaluators graded how well each team did when they submitted their results to IARPA. The teams learned in June that BlackSky, Kitware, Systems & Technology Research, and Intelligent Automation would be moving on to the second phase of the project.

The teams will have to apply their algorithms to different use cases. Cooper points out that it is too expensive and slow to design new artificial intelligence solutions from scratch. Is it possible to find construction now find crop growth? He says it is a big switch because it swaps human made changes for natural ones. Cooper says that in the third phase, the remaining competitors will try to make their work into something that could detect and monitor both natural and human made changes.

There won't be a single winner of these phrases. IARPA wants to transfer promising technology to intelligence agencies that can use it in the real world. Performance against our metrics, diversity of approaches, available funds, and the analysis of our independent test and evaluation are some of the factors that IARPA uses to make phase decisions. The best solution would be to combine parts from multiple teams. There could be no teams that make it.

Since science goes where the money goes, IARPA's investments sometimes steer scientific and technological paths. A lot of attention will be given to the problem IARPA chooses to tackle. The datasets IARPA creates for its programs, like those labeled troves of satellite imagery, are often made public for other researchers to use.

Satellite technologies are used for both military and civilian purposes. Lessons from the Kitware software will be applicable to environmental science. His company does environmental science work for organizations like the National Oceanic and Atmospheric Administration and his team has helped detect seals and sea lions in satellite imagery. He thinks of applying Kitware's software to something that already uses Landsat imagery. The rainforest in Brazil has been converted into man-made areas. "What do you think?" Hoogs asked.

The implications of auto-interpretation of landscape change are obvious for studying climate change. It is possible to spot where humans are impinging on areas of the natural landscape by looking at new construction.

The United States Geological Survey was chosen as a test and evaluation partner because of those environmental applications. IARPA's cohort is interested in the findings for their own reasons. Climate change is of great importance to the intelligence community. The second application of a dual-use technology is basically the same as the first.