AGV vs AMR: Key Differences and Which One You Need (2026)
AGV vs AMR compared — fixed-path vs autonomous navigation, flexibility, cost, safety, and infrastructure — plus how to choose for your facility.
The short answer: An AGV (Automated Guided Vehicle) follows a fixed, pre-defined path using magnetic tape, wires, or reflectors. An AMR (Autonomous Mobile Robot) navigates independently using onboard sensors and maps, choosing its own route around obstacles in real time. AGVs deliver predictability and throughput on stable routes; AMRs deliver flexibility in dynamic environments.
If that single paragraph answered your question, the comparison table below will seal the decision. If you need the full operational picture — including PLC integration, WMS connectivity, and total cost of ownership — read on.
Quick Comparison: AGV vs AMR at a Glance
| Factor | AGV | AMR |
|---|---|---|
| Navigation | Fixed path (magnetic tape, wire, laser reflectors, optical) | Autonomous (SLAM, LiDAR, camera, onboard maps) |
| Obstacle handling | Stops and waits; may alarm | Detects and reroutes dynamically |
| Infrastructure needed | Tape, embedded wire, reflectors, or floor markers | None (or minimal); map built during commissioning |
| Layout change cost | High — physical re-routing required | Low — update the software map |
| Deployment time | Weeks to months (infrastructure installation) | Days to weeks (mapping and commissioning) |
| Throughput predictability | Very high — cycle times are deterministic | Moderate — rerouting introduces variability |
| Payload capacity | Wide range; heavy-load models widely available | Generally low-to-medium (increasing in newer models) |
| Unit cost | Lower to mid-range for simple models | Mid-to-high; decreasing as technology matures |
| PLC/WMS integration | Mature, well-documented protocols | REST APIs, ROS 2, VDA 5050; newer but widely supported |
| Best fit | High-volume, fixed routes, stable layouts | Dynamic layouts, mixed traffic, frequent route changes |
What is an AGV?
An Automated Guided Vehicle is a driverless transport vehicle that moves along a predetermined path. The path is defined physically — by magnetic tape bonded to the floor, buried induction wire, retroreflective laser targets mounted to racking, or painted optical lines — and the vehicle's guidance system locks onto that physical reference to navigate.
AGVs have been a fixture in industrial manufacturing and warehousing since the 1950s. The technology is proven, the safety standards are mature (ISO 3691-4), and the PLC integration ecosystem is deep. Most AGV fleets communicate with a central fleet management controller over an industrial Ethernet network, with discrete I/O or OPC-UA handshakes to the site PLC for door control, conveyor interlocks, and charging station management.
Common AGV types include:
- Unit load carriers — transport pallets or large tote loads between fixed stations
- Tuggers and trains — tow carts along fixed milk-run routes
- Forklift AGVs — automated counterbalance or reach trucks for racking operations
- Conveyor-integrated AGVs — mate with roller conveyors at load/unload stations
For a deeper look at AGV fundamentals, see our guide on automated guided vehicles.
The defining characteristic of an AGV is its deterministic path. Every trip from Point A to Point B takes the same route, the same time, and consumes the same battery. That predictability is a strength — you can plan cycle times to the second, size buffer conveyors precisely, and synchronize the AGV with upstream or downstream PLC sequences without ambiguity.
The weakness is equally clear: change the route and you change the infrastructure.
What is an AMR?
An Autonomous Mobile Robot builds and maintains a real-time map of its environment and plans its own path through that environment on every trip. Most AMRs use SLAM (Simultaneous Localization and Mapping) — a technique that fuses LiDAR point clouds, wheel odometry, and sometimes camera data to keep the robot localized to within a few centimeters without any fixed floor infrastructure.
When an AMR detects an obstacle — a pallet left in the aisle, a pedestrian, a forklift — it recalculates a collision-free path around it and continues to its destination. If the warehouse layout changes, a supervisor remaps the affected zone in the fleet management software and pushes the updated map to all robots. No physical re-work required.
AMRs emerged commercially around 2012 and have grown rapidly. The VDA 5050 protocol (a vendor-neutral interface standard developed by German automotive associations) is becoming the dominant integration standard, allowing a single fleet management system to orchestrate AMRs from multiple manufacturers alongside AGVs.
Common AMR types include:
- Goods-to-person AMRs — carry shelving pods to picking stations (the model popularized by Amazon Robotics)
- Cart follower / tugger AMRs — autonomously tow carts on configurable routes
- Conveyor-top AMRs — carry totes between workstations in assembly and fulfillment
- Collaborative picking AMRs — travel alongside pickers to reduce walking distance
Navigation: Fixed Path vs Autonomous SLAM
This is the most fundamental difference between AGVs and AMRs, and it cascades into almost every other consideration.
How AGV navigation works
AGVs use one of several physical guidance methods:
- Magnetic tape guidance — a strip of magnetic tape bonded to the floor; the AGV's magnetic sensor array tracks the field gradient to stay on path. Inexpensive to install; vulnerable to damage from forklift traffic.
- Inductive wire guidance — a low-frequency AC current in a wire embedded in a floor groove; very robust but expensive to install and nearly impossible to re-route.
- Laser triangulation (laser nav) — the AGV spins a laser scanner and measures angles to retroreflective targets mounted on columns and racking. More flexible than tape but still requires physical target installation and surveyed coordinates.
- Optical / QR navigation — the vehicle reads floor-mounted QR markers or optical patterns. Relatively easy to re-route by repositioning markers, but still requires physical placement.
In every case, the path is encoded in physical infrastructure and reflected in the AGV's onboard route table. Deviating from the programmed route is not possible — the vehicle simply stops if it loses track.
How AMR navigation works
AMRs rely on onboard intelligence rather than infrastructure. During initial commissioning, the robot drives through the facility in "mapping mode," building a spatial map from LiDAR returns. Subsequent navigation uses a localization algorithm to match live LiDAR scans against the stored map, establishing the robot's position continuously.
Path planning runs in real time. Given a destination, the AMR calculates an obstacle-free path, executes it, and adjusts dynamically if the environment changes. Safety-rated LiDAR zones (per ISO 3691-4) provide protective field coverage in all directions of travel — the robot slows and stops without alarming if a person enters the safety field.
The operational consequence: AMRs handle the uncontrolled reality of a busy warehouse far more gracefully. AGVs require strict lane discipline from human traffic; AMRs coexist with mixed traffic by design.
Flexibility and Scalability
AGV flexibility is bounded by infrastructure. Adding a new route means installing new tape or reflectors, commissioning the route in the fleet controller, and testing. In a facility where pallet positions or conveyor stations move seasonally, this becomes expensive and operationally disruptive.
AGV fleets do scale well in controlled environments. Once the infrastructure is in place, adding vehicles to an existing route is straightforward — the fleet controller assigns missions and manages traffic at conflict points using zone-based blocking logic.
AMR flexibility is largely a software problem. New routes are defined in the fleet management UI, new zones are added to the map, and traffic rules are configured. Deploying additional robots into an existing AMR fleet is rapid: commission the robot, enroll it in the fleet, done. No floor work required.
This makes AMRs particularly well-suited to:
- Facilities that rearrange layouts seasonally (e.g., e-commerce fulfillment with campaign peaks)
- Brownfield deployments where floor modification is restricted
- Operations where multiple parallel routes need to be added incrementally
- Pilot programs where the long-term route plan is not yet fixed
For companies running high-mix, variable-volume production — the kind explored in the manufacturing automation guide — AMR flexibility often outweighs the higher per-unit cost.
Cost and ROI
Cost comparisons between AGVs and AMRs are frequently oversimplified. The unit purchase price is only part of the picture.
AGV cost factors
- Vehicle cost: Entry-level magnetic tape AGVs start in the low tens of thousands per unit. Heavy-load forklift AGVs can exceed $150,000 per vehicle.
- Infrastructure cost: Tape installation is modest ($2–$10 per meter including labor). Embedded wire or laser target surveys add significantly to the project budget.
- Maintenance: Physical guidance media degrades. Tape must be replaced when damaged or when floor coatings are refinished. Laser targets need cleaning and positional verification.
- Route change cost: Non-trivial. Tape re-routing requires facility downtime; wire re-routing is very expensive.
- Total cost of ownership: Lower than AMR on stable, long-running routes at high utilization.
AMR cost factors
- Vehicle cost: Mid-market AMRs typically range from $30,000 to $80,000 per unit depending on payload, speed, and sensor suite.
- Infrastructure cost: Minimal — commissioning labor and software licensing.
- Maintenance: Software updates, sensor calibration, battery replacement. No floor infrastructure to maintain.
- Route change cost: Near-zero for minor changes; moderate for full facility remaps.
- Fleet management software: AMR vendors typically charge annual licensing fees for their fleet management platform.
- Total cost of ownership: More competitive when layouts change frequently or when deployment speed matters.
ROI drivers for AGVs: High-volume, high-utilization, stable routes. The deterministic cycle time makes labor replacement calculations straightforward.
ROI drivers for AMRs: Dynamic environments, speed to deployment, brownfield constraints, and facilities where route infrastructure costs are prohibitive.
Obstacle Handling and Safety
Safety behavior under obstacle conditions is a major operational difference.
AGVs stop. When an AGV's safety scanner detects an obstruction in its path, the vehicle stops and waits. In most systems, it generates an alarm if the obstruction is not cleared within a timeout period. An operator must intervene. In a busy facility, a single blocked AGV can stall an entire zone if cross-traffic prevents it from clearing.
AMRs reroute. When an AMR's safety system detects an obstruction, it calculates an alternative path around the obstacle and continues — subject to available clearances in the mapped environment. If no alternative path exists (e.g., a blocked aisle with no bypass), the AMR will stop and wait, but this is a last resort rather than the default behavior.
Both AGV and AMR safety systems are certified to ISO 3691-4 and IEC 62443 (cybersecurity). The safety-rated LiDAR fields, emergency stop circuits, and bumper contacts that protect AGVs are equally present on AMRs. The difference is in the response logic, not the detection hardware.
For facilities with high pedestrian traffic — common in light manufacturing and assembly — the AMR's rerouting behavior substantially reduces operator intervention burden and improves throughput predictability under real-world conditions.
PLC and WMS Integration
Both AGV and AMR systems need to integrate with the broader automation architecture: PLCs that control doors, conveyors, and loading stations; and WMS or ERP systems that generate transport tasks.
AGV integration patterns
AGVs have a long history of PLC integration. Common patterns include:
- Discrete I/O handshake — the AGV controller exchanges 24 V DC signals with the site PLC for load-ready, unload-request, door-open/close, and conveyor-run commands. Robust, low-latency, and immune to network issues.
- OPC-UA — increasingly standard on modern AGV controllers for real-time data exchange with SCADA and MES systems.
- PROFINET / EtherNet/IP — some AGV manufacturers expose a fieldbus interface for tight PLC coupling.
- SAP/WMS REST API — the AGV fleet management software typically integrates with the WMS via REST or SOAP to receive transport orders and report completions.
The material handling PLC programming guide covers conveyor interlock and handshake patterns in detail — the same discrete I/O logic applies directly to AGV load station integration.
AMR integration patterns
AMRs have largely standardized on higher-level interfaces:
- VDA 5050 — the emerging vendor-neutral standard for AMR fleet management. Defines a JSON/MQTT message protocol for order assignment, status reporting, and action control. Enables multi-vendor fleet management.
- REST API — all major AMR vendors provide REST APIs for WMS and ERP integration. Transport orders are submitted as HTTP POST requests; status is polled or delivered via webhook.
- ROS 2 / ROS Industrial — common in research and advanced industrial deployments; provides a middleware layer for sensor fusion, path planning, and fleet coordination.
- PLC I/O handshake — retained for station-level interlocks (conveyor ready, door open) even on AMR systems.
The practical difference: AGV integration tends to be tighter at the PLC level with well-established fieldbus patterns. AMR integration tends to be looser at the PLC level but richer at the WMS/MES level, which aligns well with cloud-connected and IT-centric automation architectures.
For the broader context of how mobile robots fit into an automated production strategy, the industrial robot programming guide covers the integration stack in depth.
When to Choose an AGV
Choose an AGV when:
- Routes are fixed and unlikely to change — receiving to storage, storage to production line, finished goods to shipping. If the facility layout is stable for 5+ years, infrastructure investment is justified.
- Throughput predictability is critical — line-feeding applications where the PLC sequence depends on an exact arrival time benefit from AGV determinism. AMR rerouting introduces timing variability that can require larger buffers.
- Heavy payloads are required — forklift-class AGVs handling loads of 1,000–10,000 kg are more mature and more widely available than equivalent AMRs.
- The environment is controlled — clean, restricted-access areas with minimal pedestrian traffic are well-suited to AGV guidance. Mixed-traffic floors with uncontrolled human movement increase AGV stoppage rates.
- Long-term operating cost matters most — on a stable, high-utilization route, the lower AGV unit cost and absence of annual fleet software licensing fees can deliver a lower 10-year TCO.
- Integration with existing PLC infrastructure is required — if the facility's automation is built around a mature discrete I/O or PROFINET architecture, AGV integration is more direct.
When to Choose an AMR
Choose an AMR when:
- Layouts change frequently — seasonal SKU changes, facility expansions, lean cell rearrangements. The ability to update routes in software without floor work is a major operational advantage.
- The deployment timeline is short — AMRs can be commissioned in days versus weeks for AGV infrastructure installation. For rapid pilots or time-sensitive ramp-ups, AMR speed-to-value is compelling.
- Brownfield constraints prevent floor modification — heritage facilities with epoxy-coated floors, underfloor utilities, or listed building status may prohibit tape bonding or wire embedding.
- Mixed pedestrian and vehicle traffic is unavoidable — AMRs navigate around people naturally; AGVs require lane segregation or traffic management protocols.
- Incremental fleet scaling is needed — starting with two AMRs and growing to twenty over three years is straightforward. Scaling an AGV fleet requires proportional infrastructure expansion.
- WMS/ERP integration is primary — if the transport orchestration lives in the WMS rather than in a PLC-centric control hierarchy, AMR REST/VDA 5050 interfaces are a natural fit.
Frequently Asked Questions
What is the difference between an AGV and an AMR?
An AGV follows a fixed, physically defined path (magnetic tape, embedded wire, or laser reflectors) and stops when it encounters an obstacle. An AMR navigates autonomously using onboard sensors and maps, reroutes around obstacles in real time, and requires no floor infrastructure. The core distinction is navigation method: infrastructure-guided vs onboard intelligence.
Is an AMR better than an AGV?
Neither is universally better. AMRs are better for dynamic, frequently changing environments where flexibility and fast deployment matter. AGVs are better for stable, high-volume routes where deterministic cycle times and lower infrastructure-free unit cost are priorities. The right choice depends on your specific operation.
Which is cheaper, an AGV or an AMR?
Entry-level AGVs can be less expensive per unit than comparable AMRs, but total project cost must include infrastructure installation, which can equal or exceed the vehicle cost. AMRs have negligible infrastructure cost but typically carry higher per-unit prices and ongoing fleet software licensing fees. On stable, long-running routes at high utilization, AGVs often win on TCO. In dynamic environments, AMRs win.
When should you use an AGV instead of an AMR?
Use an AGV when your routes are fixed and stable for the long term, when you need heavy-lift capacity, when throughput predictability is non-negotiable for PLC-synchronized line feeding, or when the environment is controlled with minimal pedestrian traffic. The deterministic cycle time of an AGV is an operational asset in these scenarios, not merely a limitation.
Summary
AGVs and AMRs solve the same problem — autonomous material transport — with fundamentally different approaches. AGVs trade flexibility for determinism. AMRs trade determinism for adaptability. Neither is the right answer in every situation.
For operations teams planning an intralogistics automation project, the decision framework is straightforward: map your route stability, your layout change frequency, your traffic mix, your payload requirements, and your integration architecture. The answers will point clearly toward one technology or the other — and in large facilities, often toward a hybrid fleet running both.
As autonomous mobile robot technology matures and AGV navigation systems incorporate more onboard intelligence, the line between the two categories is narrowing. For now, understanding the operational implications of each approach is the foundation of a successful automation investment.
