What Is an AMR?
Autonomous mobile robots (AMRs) are devices that can sense, think and move autonomously. They often work independently to ferry materials across warehouses, factories, hospitals and even public sidewalks, but they also help in activities like security and data collection.
AMRs build on the earlier generation of mobile robotics, referred to as automated guided vehicles (AGVs). AGVs follow fixed routes — much like trains on tracks — guided by magnetic strips, QR codes or predefined paths. They excel in predictable environments where material flows never change. If an obstacle appears, an AGV simply stops and waits. Its movements are automated, and it does not interpret or adapt to its surroundings.
AMRs, in contrast, are built for autonomy. Using combinations of sensors, perception systems and onboard computing, they navigate through unstructured indoor and outdoor environments without needing to follow predefined tracks. AMRs can reroute when a pallet is blocking an aisle, slow down as a person passes in front of them, or select an entirely new path when a warehouse layout changes. They come in many forms: squat platforms that slide beneath a pallet and lift it, autonomous pallet jacks and forklifts, compact delivery robots designed for sidewalks, and even quadrupeds, such as Spot, Boston Dynamics’s famous dog-bot. The common thread is mobility powered by sensor-driven intelligence that is sometimes reliant on the same technology as ADAS-equipped vehicles. With their mobility and autonomy, AMRs are prime examples of physical AI’s “sense, think, act and optimize” process.
Similar to the stationary robotic arms called “cobots” (collaborative robots), AMRs can safely operate around humans. However, the robotics industry draws a clear line between those categories. Cobots are manipulation systems; AMRs are mobility systems. When a robotic arm is mounted on a mobile base, the industry classifies the unit as a “mobility manipulator” rather than a cobot, because its defining function is movement. Several major robotics manufacturers reinforce this separation by maintaining distinct product categories for “industrial and collaborative robots” versus AMRs. This distinction matters because it governs not only the types of tasks these machines perform but also expectations around safety, capabilities and regulation.
To develop the perception systems necessary for AMR autonomy, many AMR manufacturers rely on expensive arrays of lidars, laser scanners and stereo cameras simply because these combinations have become industry‑accepted, not because they are cost-optimized. That reliance has contributed to high system costs and slower adoption, revealing a clear opportunity for innovation.
Ultimately, cobots, AGVs and AMRs address different needs within the broader automation ecosystem. Cobots serve as precision assistants, augmenting human workers at fixed stations. AGVs provide basic point‑to‑point transportation in structured environments. AMRs add intelligence and flexibility, offering autonomous movement in dynamic spaces. While future systems may blur these boundaries — particularly as mobile robots increasingly incorporate manipulation capabilities — today the distinction remains central to how the robotics industry defines, develops and deploys these technologies.