Reality does 3D mapping and exploration.
Built in c++, it has algorithms for edge/pattern/line/noise detection/mapping/comparisons.
Thereafter it shares the processed images of the 2 camera's over the 2 SBC's, calculating a disparity map and the trajectory the camera's have done. A lot of complicated translations of pixels to real world info. With a 3D map, path finding isn't that hard anymore.
It has a scheduler and (literal) task manager for orderly execution of work. In the future I'll make it able to learn specific acts you can bind to words.
The robot itself can lift 500g with a stretched arm, has 2 so lifts 1kg and maybe 2kg with adjusted angles.
16 Lithium Ion cells, averaging 3000Mah each. 100+ hours of standby, 10+ hours of driving, 2 hours of arm motion.
It currently drives at 4 km/h with good torque, functional to not have it crash into a wall so later I might upgrade.
