Communicating with robotic assets in space and on the surfaces of distant planets is tricky. The great distances any communication needs to travel means that any command given can take between several minutes and many hours to be received by a robotic asset. Once received, it then takes equally as long for operators to receive confirmation either via telemetry of via a video feed that the commands have been executed successfully.
NASA hopes to negate much of the trouble caused by communication delays by utilising AI to handle a number of the basic operations of a particular robot’s duties. An operator would simply need to send a list of actions he requires to be completed. The onboard AI would then calculate how best to approach each task and then execute them in the most effective order.
This kind of AI implementation is, however, just the start. NASA engineers imagine scenarios in which an AI-piloted interstellar spacecraft could select which planets to orbit and how best to go about documenting the planet. It could then pick the best locations to send probes down to the surface and monitor their progress. In short, an entire mission itinerary could be created and executed by the spacecraft without any human intervention whatsoever. The information collected by the autonomous craft could then be used to select the most interesting locations for manned missions.
Although the space agency’s ambitions may sound more like science fiction than science fact, NASA has already started using rudimentary AI in its missions. The Curiosity Rover, for example, utilises automated targeting systems that dictate which rocks the rover’s cameras and sampling lasers examine. In essence, it evaluates the terrain and decides what’s most important to inspect. The technology aboard Curiosity is the latest generation of self-targeting systems that were originally developed for the Opportunity and Spirit rovers. Opportunity has impressively been running on the surface of Mars for over 13 years performing much of its day to day operations autonomously.
Additionally, autonomous spacecraft have seen operations much closer to home. The Autonomous Sciencecraft Experiment (ASE), for instance, has just completed its mission after being launched back in 2003. ASE was tasked with observing Earth’s surface to look for anything interesting like volcano eruptions or severe weather patterns. It is hoped that the lessons learned in the ASE mission will help create next generation AI monitoring satellites that will be able to find and alert the public to natural disasters.
AI technology is not without its detractors with even Elon Musk famously voicing concern at the speed at which these programmes are being allowed to develop. However, with the vastness of space being much more than the human race could possibly hope to explore within our extensive, the prospect of autonomous exploration is just too great an opportunity to pass up.