Not so distant future. A colony of autonomous space drones arrives at an asteroid belt. The colony ship contains construction facilities, a basic refinery and a few mining drones. Everything is of most basic, rugged design: hydrogen based propulsion, solar for energy, steel frames, silicone and copper based electronics, standardized, easy to manufacture in space components. All controlled by an AI.
The purpose of this colony is to mine resources out of the belt to eventually build a megastructure in the system: a dyson sphere, a hyperspace gate, a mass relay, whatever. The construction requires truly massive amounts of resources and so you need to build an equivalently big infrastructure to collect them. Worst of all, you're not alone in the belt. Other colonies are arriving and have similar initiatives. Each AI is independent and has its own opinion on where the path of least resistance lies. AIs may decide to work together on the same project, keep neutrality and self-regulate, aggressively compete between each other or steal each other resources. There are no rules, there's only the goal.
The colony arrives at a random place inside the belt. It can later reposition itself, but it requires propulsion mass and there's very little of it left at the beginning. The colony sends basic drones to the nearest asteroids for mining. Basic drones can break small chunks of the asteroid and either carry or hurl them towards the colony. The colony then refines the chunks into resources to fuel existing drones and build new ones.
Space physics apply to every kind of motion in here. Movement from point A to point B is either acceleration, deceleration or drift. Your travel time depends on how much time you spend accelerating\decelerating comparing to drifting. The heavier you are and the faster you want to travel, the more fuel you have to burn. The more fuel you carry, the heavier you are. It might end up being more economical to hurl the rocks you just mined towards your colony and catch them on the other end than to carry them back all the way. It might be even more beneficial to invest into a smaller refinery and park it closer to asteroid to only send back to colony resources that are already refined and are lighter. It might be worth moving the entire colony closer to the asteroids or move whole asteroid closer to the colony.
Another side effect of space physics is that change of destination for a squad that is already in flight may be very slow and costly. If you're going full thrust, approaching midpoint for a flip and burn and then decide to turn back, well, you're shit out of luck, you might as well finish your trip before turning back. Or you can burn twice as hard the other way, providing you even have enough propulsion mass for the maneuver. And if you do, why the hell do you carry that much extra weight?
The drones usually work in squads. The composition of the squad determines it's capabilities. Typical classes of drones include:
- miners - drones equipped with short range energy weapon mainly used to break chunks of asteroids, basic propulsion and power system. This are more generic kind of drones. Jack of all traits, master of none, as they say.
- security - drones lack propulsion and have barely sufficient power system, but have PDCs that could destroy other drones in the vicinity
- refineries - process rocks into resources. No propulsion or power.
- construction - build other drones from resources. No propulsion or power.
- propulsion - logistic drones. Can effectively transport other drones and resources around the belt.
- power - drones providing power to other drones. Lack of power may reduce the rate at which security and refinery drones operate.
- attack - drones specifically designed to destroy defenses and clear the way to other types of drones
- and many more.