SLAM vs Path Planning
People think SLAM finds the way to the goal, and it does not. SLAM answers two questions: what is around me, and where am I on that map. Path planning takes the map and answers a third: how do I get from here to there without hitting anything. A robot needs both, and SLAM runs first.
Simultaneous Localization and Mapping (SLAM)
A robot draws a map of a new place and tracks where it is on it, at once.
SLAM is how a robot maps a new place and tracks where it is on that map. The two jobs depend on each other. To find itself it needs a map, and to draw the map it needs to know where it is. So it solves both at once, step by step, with cameras, lidar and how far its wheels have turned. Each time it sees a place it knows, it fixes small errors in both. Robot vacuums, drones and warehouse carts all use it.
Path Planning
Working out a safe route for a robot. It goes from here to the goal.
Path planning is how a robot works out a safe route from where it is to where it needs to be. It starts with a map of what is free and what is blocked. It then searches for a route that misses every obstacle. The best route is often the shortest, but it may also need to be smooth or keep a safe gap. The robot plans again whenever something new gets in the way.
In one line
- Simultaneous Localization and Mapping (SLAM)
- A robot draws a map of a new place and tracks where it is on it, at once.
- Path Planning
- Working out a safe route for a robot. It goes from here to the goal.