What is the main difference between an AGV and an AMR?

Navigation. An AGV follows a fixed physical guide path, magnetic tape, wire, QR codes, or reflectors, and stops when blocked. An AMR uses onboard sensors and SLAM to build its own map, locate itself, and route around obstacles in real time. AGVs are told where to go by the building; AMRs figure it out.

What does SLAM mean for mobile robots?

SLAM stands for simultaneous localization and mapping. An AMR uses LiDAR and cameras to build a digital map of its environment while simultaneously tracking its own position within that map. This lets it navigate and avoid obstacles without fixed floor guides, and adapt when the layout or traffic changes.

Are AMRs replacing AGVs?

AMRs are growing faster and displacing AGVs in dynamic, frequently-changing environments where flexibility matters. But AGVs remain the better choice for stable, high-volume, fixed routes, where their simpler navigation is cheaper and more reliable. Many plants run both, AGVs on fixed trunk routes and AMRs on variable last-stretch moves.

Which is cheaper, an AGV or an AMR?

It depends on what you count. AGVs often cost less per vehicle but more up front because they require floor infrastructure (tape, wire, reflectors) and are expensive to re-route. AMRs usually cost more per vehicle but need no floor changes and reroute in software, so total cost depends heavily on how often your routes change.

Do I need fleet management software for mobile robots?

Once you have more than a couple of vehicles, yes. Fleet software handles traffic sequencing and deadlock avoidance, assigns tasks to the best-positioned vehicle, orchestrates charging to keep the fleet available, and integrates with your WMS, MES, or ERP so move requests become actual vehicle motion and completion flows back.

Can AGVs and AMRs connect to my MES or ERP?

Yes, and they should. The value of an automated fleet comes from integration: move requests originate in your systems of record and completion flows back, so material moves in response to what the plant actually needs. A fleet disconnected from those systems is just expensive choreography moving loads blindly.

AGVs vs AMRs: Automated Material Movement

Q: What safety standards apply to AGVs and AMRs?

ISO 3691-4:2023 covers driverless industrial trucks and applies to most AGVs and many AMRs. ANSI/RIA R15.08, developed by A3 in the U.S., is written specifically around autonomously navigating industrial mobile robots. Any deployment needs a risk assessment covering speed, stopping distance, pedestrian interaction, and emergency stops.

Fixed-path AGVs and free-roaming AMRs both move material without a driver, but they navigate, adapt, and fail in very different ways, and the right choice depends on your floor, not the brochure.

An AGV (automated guided vehicle) follows a fixed path laid into the floor, magnetic tape, buried wire, QR codes, or reflectors, and halts when the path is blocked. An AMR (autonomous mobile robot) uses onboard sensors and SLAM to map its surroundings, locate itself, and route around obstacles on its own. Both move material without a driver. How they navigate is the whole difference.

The two acronyms get used interchangeably, and vendors are not always in a hurry to correct it. But the distinction is not marketing, it decides how much you spend on floor infrastructure, how the system behaves when a pallet is left in the aisle, and how painful it is to change a route six months from now. This post separates AGVs from AMRs on navigation, flexibility, cost, and safety, walks through where each one actually fits, and gives you a way to choose. If you are automating material flow more broadly, read it alongside packaging line automation which often sits at the ends of these routes.

What is the difference between an AGV and an AMR?

The difference is how the vehicle knows where to go. An AGV is told where to go by the building: it tracks a physical guide path installed in or on the floor and follows it like a train on rails. An AMR is not told, it figures it out, carrying its own map and sensors and planning a route the way a person crossing a warehouse would.

That single design choice ripples into everything else:

Navigation. AGVs follow magnetic strips, inductive wire, optical lines, QR-code grids, or laser reflectors. AMRs run SLAM simultaneous localization and mapping, using LiDAR and cameras to build and update a map on the fly.
Obstacle handling. Block an AGV's path and it stops and waits, because it cannot leave the line. Block an AMR and it plans a path around the obstacle and keeps moving.
Changing a route. Rerouting an AGV usually means relaying tape or wire and reconfiguring the path, physical work. Rerouting an AMR is a software update to its map.
Environment fit. AGVs want stable, predictable, high-traffic-controlled space. AMRs are built for dynamic, cluttered, frequently-changing floors.

The defining behavioral difference. An AGV bound to its guide path stops at an obstacle; an AMR carrying its own map plans a new route around it and continues.

An AMR localizes itself against a map it builds and maintains with SLAM. On its first pass it drives the space and constructs a digital map from LiDAR range data and camera imagery. From then on, it continuously compares what its sensors see to that map to answer two questions many times a second: where am I? and what changed? When a forklift, a person, or a stray pallet appears, the AMR treats it as a dynamic obstacle, replans locally, and keeps its task moving.

AGV navigation is deliberately simpler and, in the right setting, more robust. A wire or magnetic tape gives an unambiguous path with no perception required; the vehicle just follows the signal. That simplicity is a feature where routes never change and traffic is controlled, it is cheaper to run and easier to trust. It becomes a liability the moment the floor is dynamic, because the AGV has no way to improvise.

When should you choose an AGV over an AMR?

Choose the AGV when the route is fixed, the volume is high, and the environment is stable; choose the AMR when the routes vary, the layout changes, or the floor is shared with people and forklifts. Neither is universally better, they optimize for opposite conditions.

Factor	AGV favors	AMR favors
Route stability	Fixed, repetitive, unchanging	Variable, frequently re-planned
Layout changes	Rare (relaying guides is costly)	Common (map update in software)
Environment	Controlled, low unexpected traffic	Dynamic, shared with people/forklifts
Volume & predictability	High-volume, steady flow	Mixed loads, changing demand
Up-front infrastructure	Higher (floor guides, install)	Lower (no floor changes)
Per-unit vehicle cost	Often lower per vehicle	Often higher per vehicle

AGVs and AMRs optimize for opposite conditions. The decision is rarely about which technology is newer and almost always about how stable your routes and floor are.

A useful rule of thumb: if you could draw the route on the floor with tape and it would still be correct in two years, an AGV is probably the cheaper, more reliable answer. If the route depends on today's orders, today's layout, and who is parked where, the AMR's autonomy is worth its premium. Many plants end up with both, AGVs on the trunk routes, AMRs handling the variable last hundred feet.

What does fleet management add?

Once you have more than a couple of vehicles, the fleet manager becomes the real system. A single robot moving a load is a demo; a coordinated fleet keeping a plant supplied is an operation. Fleet software does the work no individual vehicle can:

Traffic management. It sequences vehicles through intersections and narrow aisles so they do not deadlock or collide, the way a controller sequences trains.
Task allocation. It assigns each move to the best-positioned vehicle and rebalances when demand shifts, so the fleet behaves like one system rather than a crowd.
Charging orchestration. It cycles vehicles to chargers to keep enough of the fleet available without starving the routes, an availability problem that looks a lot like machine monitoring on wheels.
Integration. It talks to the WMS, MES, or ERP so a move request from the system of record becomes a vehicle actually moving a load, and completion flows back.

Fleet management turns individual vehicles into one coordinated system: it takes move requests from the systems of record and handles traffic, task allocation, and charging so the fleet behaves as a whole.

What safety standards apply to AGVs and AMRs?

Both fall under mobile-robot safety standards, and treating them as ordinary machines is a mistake. The two references to know:

ISO 3691-4:2023 Driverless industrial trucks and their systems. It grew out of the AGV world and specifies safety requirements for driverless trucks; most AGVs and many AMRs are covered by it.
ANSI/RIA R15.08 the U.S. industrial mobile robot safety standard developed by A3, written from the start around autonomous navigation with no assumption of a guide path.

The practical upshot: any deployment needs a risk assessment covering speed, load, stopping distance, pedestrian interaction, and emergency stops, and the newer AMR standard exists precisely because autonomous routing raises safety questions that the guide-path world never had to answer.

How do you decide what to deploy?

Work from the flow, not the technology. A short, disciplined evaluation keeps you from buying a fleet to solve a problem a conveyor would have handled:

Map the actual moves. Which loads, from where to where, how often, at what times. Automate flows that are real and repetitive, not the ones that are merely annoying.
Judge route stability. If the routes are fixed for years, lean AGV. If they shift with orders or layout, lean AMR. Be honest about how often your floor really changes.
Assess the environment. Shared aisles, pedestrians, forklifts, and clutter push toward AMR autonomy; controlled, low-traffic lanes make AGVs safe and cheap.
Total the real cost. Add floor infrastructure, install, fleet software, integration, and charging to the vehicle price. AGVs often cost less per vehicle but more to install and re-route.
Plan the integration. Decide up front how move requests reach the fleet from your WMS, MES, or ERP, and how completion flows back, this is where projects stall.
Pilot one flow, then scale. Prove one route end to end, measure it against the manual baseline, and expand only once the fleet earns it.
Design safety in from day one. Run the ISO 3691-4 / R15.08 risk assessment before deployment, not after the first near-miss.

Where do AGVs and AMRs fit in a connected plant?

They are one more source of real-time operational data, and they are only as valuable as their integration. A fleet moving material blindly, disconnected from the systems that know what the plant needs, is expensive choreography. Connected to the systems of record and to live floor status, it becomes responsive material flow. Industry analyses put the autonomous mobile robot market in the low-single-digit billions of dollars in 2026 and growing at a double-digit CAGR, driven by exactly this pull, flexible material movement in warehouses and plants facing labor constraints.

The value shows up when vehicle status, task completion, and material location land in the same live picture as machines, quality, and schedule, the substance behind smart factory technology the Industrial IoT and the digital twin. That is what Harmony builds toward: one real-time view that connects machines, systems, and now mobile fleets across a plant without ripping out what you already run. See how a connected floor looked in practice in the CLS case study or walk the module map on the features section of our homepage.