The National Aeronautics and Space Administration (NASA) has announced a project where it is inviting everyone to help in programming its Mars rover, Perseverance. The project is called AI4Mars, and simply put, the simple goal is to speed up the ability for NASA operators to instruct its rover better on the Red Planet. The crowdsourcing project is now live, and is designed as a way for space enthusiasts and self-proclaimed explorers to feel closer to the agency’s space missions.
AI4Mars is solely focused on improving the ability for NASA ground operators to navigate the kind of data that its Perseverance rover returns on a daily basis. This daily data dump includes up to hundreds of images of the Martian terrain, which scientists and operators then use to understand the planet better. This, in turn, also allows them to give clearer instructions to the AI algorithm on board Perseverance, which then uses the fed-in data to better gauge the Martian surface in future explorations.
With AI4Mars, NASA is now uploading a wide variety of the images returned by the rover to give viewers a chance to help out. Once a viewer sees it, they can then help add labels to the Martian soil. This exercise can make significant difference in terms of not just crowdsourcing expertise, but also in identifying terrain patterns that even NASA’s operators may have missed — or simply taken too long to spot. Viewers can add granular surface details through their labels as well, hence helping Perseverance better identify terrain alterations.
The programme is interestingly based on a tool that the previous Mars mission by NASA, dubbed Curiosity, gave birth to. The tool set is called ‘Soil Property and Object Classification’ (SPOC), and came into being with Curiosity in order to identify the Martian terrain and help operators navigate it significantly better than before.