This year's ICFP Contest might help bring space-related AI challenges to the the attention of the programming languages community. The task of the contest is to deploy control software for a Martian lander and steer a rover from its starting position to its home base.
While the details of the task are abstracted from real deployment scenarios or even tongue-in-cheek, the fundamental idea is a point of very much debate and research. Increased autonomy of spacecraft and rovers is very much a multi-dimensional optimisation challenge in terms of science return and safety.
Communication with spacecraft increases in cost the further away the object is from Earth. On the other hand, the science instruments may detect interesting events that require timely action. The event might not be over before decision makers in mission control are able to update the schedule.
This is where increased autonomy might come into play. Of course, the safety of the mission must not be compromised by having the spacecraft issue commands that impact its own functioning. But science return might be increased significantly by identifying simply cases where the vehicle is allowed to deviate from its given schedule and return to normal operation after ad-hoc observation that were triggered by certain events.
Ideally, software tools should be adapted to quantify the benefits, impacts and drawbacks of such ideas. White Label Space hopes that the passionate participants of contests like this year's ICFP Contest help enlarge the pool of technologies from which the space community can draw their technologies.