The skies above us are not just empty space; they are teeming with thousands of satellites, including those designed to observe our planet. Over 8,000 active satellites currently orbit Earth, with more than a thousand focused on Earth observation.
In this ever-evolving landscape, the convergence of satellite technology and artificial intelligence (AI) is poised to revolutionize how we use satellite imagery and who can access it.
Satellites: From cold war secrets to everyday tools
In the early days of satellite technology, during the Cold War era, launching satellites marked a country’s prowess in rocketry and provided crucial surveillance capabilities. These satellites were largely the domain of governments, but today, private enterprises have joined the space race, deploying satellites for various purposes, from internet coverage to Earth observation.
Among the key challenges in the satellite industry has been the analysis and interpretation of the vast amounts of data collected. New AI tools, such as Meta’s Segment Anything Model, are now proving effective in identifying objects within satellite images, streamlining the process of information extraction.
One of the most significant breakthroughs in satellite imagery utilization is the integration of large language models like OpenAI’s ChatGPT. In collaboration with Microsoft, companies like Planet Labs aim to create a “queryable Earth,” enabling individuals to interact with Earth’s surface data like data scientists query databases.
This democratization of satellite intelligence, once the preserve of classified government agencies and those with substantial resources, is poised to become accessible to anyone with an internet connection.
A brief history of satellite reconnaissance
The use of satellites for surveillance and intelligence gathering dates back to the early days of the Cold War. In response to the Soviet Union’s launch of Sputnik 1 in 1957, President Dwight D. Eisenhower authorized the Corona program to develop satellite reconnaissance capabilities.
By 1960, the United States had received its first satellite image of Soviet airfields. The subsequent decades witnessed the growth of satellite surveillance, primarily focused on monitoring enemy capabilities and ensuring compliance with treaties, including satellite surveillance provisions.
Satellite technology continued to advance rapidly, transitioning from film-based data collection to real-time transmission. Initially designed for geological observation, Landsat satellites played a crucial role in this evolution.
By 1972, they were transmitting multispectral data to Earth, significantly enhancing the capabilities of surveillance. However, this wealth of data remained classified throughout the Cold War.
Commercialization and democratization
The turning point came in 1992 when Congress passed the Land Remote Sensing Policy Act, allowing commercial companies to operate satellites and sell their data. This legislative change marked the birth of the commercial space industry, driven by advances in computing power, internet connectivity, and increased demand for services.
Despite the rapid growth of the commercial satellite industry, some aspects remained under government control due to national security concerns. Satellites utilizing Synthetic-Aperture Radar (SAR), which can produce clear images even under adverse conditions, were long restricted.
However, this changed in 2015 when XpressSAR received the first commercial license to operate SAR satellites in the United States. International demand for SAR imaging capabilities has led to other countries, such as Finland’s ICEYE, supplying these technologies for various applications, including conflict monitoring.
Commercial satellite operators have expanded their customer base beyond federal contracts. Today, customers range from investment firms monitoring industrial activity to agricultural companies assessing crop health and mining companies tracking changes in elevation. The satellite industry continues to grow as the cost of launching satellites into space decreases.
The true potential of satellite data lies in its interpretation. Recent advancements in machine learning, particularly AI, have streamlined satellite data analysis.
Tools like Meta’s Segment Anything Model and large language models like ChatGPT enable more efficient and accurate extraction of information from satellite imagery. This transformation is making satellite data more accessible and valuable than ever before.
