proprioceptive vs exteroceptive sensing
Robots learn about the world through two very different kinds of sensing, and the names sound fancier than the idea behind them. Proprioceptive sensing is when a robot senses its own body — where its joints are bent, how fast a wheel is spinning, how much its battery is drained. Exteroceptive sensing is when a robot senses everything outside its body — a wall ahead, a person walking by, how bright the room is. The simplest way to remember it: proprioceptive means inside-looking, exteroceptive means outside-looking.
You have both senses too. Close your eyes and touch your nose with a fingertip — you can still do it, because your body knows where your own arm is without looking. That inner sense of your own posture is proprioception. Now open your eyes and spot a cup across the table — that is exteroception, sensing the world out there. A robot's joint angle sensors and wheel counters are its inner sense; its cameras, microphones, and distance scanners are its outward sense.
Robots need both because each covers the other's blind spot. Inner sensing is fast, cheap, and reliable, but it cannot tell the robot that the floor is icy or that someone moved the furniture. Outer sensing reveals the changing world, but it is noisier and easily fooled by glare, fog, or shadows. Good robots blend the two: the inner sense keeps the body coordinated, while the outer sense keeps the robot aware of where it actually is and what is around it.
A delivery robot uses its wheel encoders (proprioceptive) to know it has rolled forward two meters, and its laser scanner (exteroceptive) to notice a closed door it must now route around.
Inner sense tracks the robot's own motion; outer sense reacts to a world that keeps changing.
The line can blur: a camera that watches the robot's own gripper is doing exteroceptive sensing pointed inward, so think of the two as a spectrum, not strict boxes.