Autonomous mobile robot (AMR)

An autonomous mobile robot, commonly abbreviated AMR, is a robot that moves through its environment on its own terms, using onboard sensors and software to perceive its surroundings, build a map, plan routes, and avoid obstacles as it goes. The defining feature is flexibility. An AMR is not tied to a fixed path, so if a route is blocked it can find another way, and if the layout of a space changes it can adapt without physical modifications to the environment. This makes AMRs well suited to dynamic settings where people, objects, and tasks are constantly shifting.

The usual point of comparison is the automated guided vehicle, or AGV, which represents the older approach. An AGV follows predefined routes marked out by wires, magnetic strips, or floor markers, and if its path is obstructed it simply stops until the way is clear. AMRs replace that rigid guidance with sensing and intelligence, navigating by understanding the space rather than following a track. This distinction, flexible autonomous navigation versus fixed-path guidance, is the key thing that separates the two and explains why AMRs are increasingly favored for adaptable material handling.

An AMR perceives its surroundings with sensors such as LiDAR and cameras, uses that input to build and update a map of the space, and localizes itself within that map. Given a goal, it plans a route and follows it while continuously watching for obstacles, replanning on the fly when something blocks the way. Because it relies on understanding the environment rather than on physical guides, it can operate in spaces shared with people and adapt as those spaces change. This combination of mapping, localization, planning, and obstacle avoidance is what gives an AMR its autonomy.

AMRs matter because they bring adaptable, autonomous movement to environments that used to require rigid automation or manual labor, which is a significant shift in logistics and manufacturing. For anyone tracking applied physical AI, they are one of the most commercially established categories of mobile robot, and they illustrate how sensing and software can replace fixed infrastructure. Understanding the AMR-versus-AGV distinction clarifies a common source of confusion and highlights why flexible navigation is such a valued capability.

An AMR navigates flexibly using sensors and software, choosing its own paths and rerouting around obstacles, while an AGV follows fixed routes marked by physical guides and stops when blocked. The AMR understands its environment, whereas the AGV follows a predefined track.

They are widely used in warehouses, factories, and logistics operations, where they move goods through spaces that change frequently and are often shared with people. Their ability to adapt to changing layouts is a key reason they are chosen for these settings.

It uses onboard sensors to perceive and map its surroundings, localizes itself within that map, plans routes to its goals, and avoids obstacles in real time, replanning when necessary. This lets it move autonomously without fixed guidance infrastructure.

Humanoid robot, Occupancy grid, Sensor fusion, Physical AI, Robot Operating System (ROS).