Space – A natural field of use for AI
A substantial number of AI applications are being deployed in the space and satellite industry. This is a natural field to use AI, given the very limited current ability to maintain human presence in space and the need for autonomous operations, as well as the space domain being a source of extensive raw data. Many terrestrial AI applications required AI to manage matters that humans cannot address effectively. In space, many of the operations that are in need of automation are done without any direct human presence or even human involvement.
Exciting new applications are occurring in numerous areas, including:
- Space robotics, particularly for In-Space Servicing, Assembly, and Manufacturing (ISAM)
- Avoiding collisions and monitoring of space debris
- Satellites connecting in space (rendezvous, proximity operations, and docking, or RPOD)
- Space exploration including moon and Mars rovers
- Many analytics applications, as huge amounts of data are being gathered from space
Data sensed from space (from visual to radar-based) is a seemingly endless source of new data, and machine learning algorithms are being used to process satellite imagery, detecting and classifying Earth’s features for geographical information systems, classifying various land cover types in imagery, crop monitoring and predicting, wildlife conservation, disaster response, and in connection with many other use cases.
As with terrestrial AI, the technical challenges to creating and deploying space-based AI are considerable. However, every year more progress is made at being able to control space activities both from the ground and in space itself, at collecting data in space that cannot be obtained from the ground and in furthering space exploration. Having privately-owned, manned stations in space seems within reach, and productive trips to Mars or even asteroids could well happen in the next decade.
Read more of our analysis here.