Mobile and logistics robotics

AGV Robots

Automated guided vehicles that follow defined paths for material movement.

Quick decision summary

What to know before reading the full guide

Plain definition

An AGV is a powered, driverless industrial vehicle that transports loads over defined routes. Guidance can use embedded wire, magnetic tape or spots, optical markers, reflector-based laser localization or natural features. Modern products blur the boundary with autonomous mobile robots, so buyers should verify actual route and recovery behavior instead of relying on the label.

Best-fit work

warehouse transport; hospital logistics; factory line supply

Main deployment risk

Performance drops when sensors face glare, dust, occlusion, deformable objects, poor lighting, water, smoke, or unexpected human behavior.

Measure in a pilot

completed loaded missions per hour and per shift, on-time pickup and delivery rate, manual interventions per 100 missions, blocked-route and safe-stop minutes

Research brief

Updated August 12, 2026

Why this robot category matters

An automated guided vehicle is useful when material must move between known pickup and drop-off points with repeatable timing. The vehicle is only one part of the installation. Route design, load-transfer hardware, fleet control, charging, traffic rules, safety validation, maintenance and exception recovery determine whether the system produces reliable transport or simply creates a new queue in the aisle.

The hard deployment questions are concrete: Can the vehicle maintain the required missions per hour at peak traffic? What happens when a pallet is misaligned, an aisle is blocked, localization confidence drops or a door controller fails? A credible pilot measures completed loaded missions, interventions, safe stops, queue time, charge time and recovery time under the proposed payload and floor conditions.

Buyer decision guide

AGV or AMR: choose from the workflow, not the label

AGV and AMR terminology overlaps across vendors. Compare route behavior, recovery, traffic control and operating limits in the proposed facility. A product label does not establish safety or prove throughput.

Practical comparison of AGV and AMR deployment characteristics
Decision point	Traditional AGV	AMR pattern
Route behavior	Follows a defined guide path or permitted route network.	Plans through a mapped operating area and can replan within its rules.
Blocked aisle	Usually slows or stops until the route clears or recovery is requested.	May calculate another safe route when the map and vehicle capability allow it.
Facility changes	Guide paths may need engineering and revalidation.	Maps, zones and missions still need updating, testing and validation.
Strongest fit	Stable, repetitive flows with predictable handoff points.	Changing routes, variable destinations and busier shared spaces.

Five measurements for a useful pilot

Flow

Loaded missions per hour, empty travel, route length, peak queues and handoff time.

Vehicle

Rated payload including the top module, turning envelope, gradients, floor joints, doors and lifts.

Safety

Stopping distance at rated load and speed, protective fields, emergency stops and operating-zone risk assessment.

Operations

Charge strategy, recovery owner, spare coverage, service response and restart procedure.

Integration

WMS, MES, PLC, conveyor, door and elevator interfaces; traffic control; supported VDA 5050 version.

Use cost per completed loaded mission

Annualize the vehicle, integration and infrastructure costs; add service, energy and recovery labor; then divide by completed loaded missions. Keep blocked and manually recovered missions visible.

cost per useful mission = annual operating cost / completed loaded missions

What it is

How it works

A fleet or plant-control system releases a transport mission and reserves the required route. The vehicle localizes against its guide path or map, closes a motion-control loop with wheel encoders and steering feedback, and uses safety-rated scanners or other protective devices to slow or stop near people and obstacles. A lift, conveyor, fork, pin or tugger interface transfers the load. The mission ends only after the destination handshake is confirmed; charging, faults and blocked-route recovery remain part of the operating cycle.

System architecture

01Vehicle platform sized for the complete load, top module, turning envelope, floor condition and required stopping performance.

02Guidance and localization using wire, tape, magnetic spots, optical markers, reflectors, natural features or a validated combination.

03Safety control that evaluates protective fields, speed, braking, emergency stops and vehicle state independently from ordinary mission logic.

04Fleet control that releases orders, reserves segments, manages intersections, prevents deadlocks and schedules charging.

05Load-transfer interface coordinated with conveyors, racks, doors, elevators, PLCs or manual handoff stations.

06Operations layer for alarms, logs, intervention workflow, maintenance, spare coverage and performance reporting.

Perception layer

01Safety laser scanners or equivalent protective devices monitor defined fields around the moving vehicle.

02Guidance sensors detect the installed route reference or natural features used for localization.

03Wheel encoders, steering feedback and an IMU estimate motion between external references.

04Load, fork-height, mast, docking or presence sensors confirm that pickup and drop-off actions completed correctly.

Localization and mapping

01Odometry from wheels, joints, inertial sensors, visual motion, acoustic sensing, or external references.

02Maps may represent geometry, semantic objects, safety zones, inspection assets, crop rows, racks, or work cells.

03Robust systems detect when maps are stale or localization confidence is low.

04Fallback behavior is critical because a robot using the wrong map can become unsafe.

Actuation and control

01Motion control for differential drive wheels, mecanum wheels, steered wheel modules, electric traction motors.

02Trajectory tracking with speed, acceleration, force, thermal, collision, and payload constraints.

03Recovery behaviors such as retry, reverse, replan, slow down, dock, ask for help, or safe stop.

04Human override and audit logs so operators can understand failures.

Hardware stack

01Common sensors: 2D LiDAR, 3D LiDAR, RGB cameras, depth cameras, IMU, wheel encoders.

02Movement and tools: differential drive wheels, mecanum wheels, steered wheel modules, electric traction motors, braking systems.

03Compute: embedded CPUs, GPUs, microcontrollers, motor drivers, safety controllers, and networking.

04Mechanical design: stiffness, cable routing, ingress protection, cooling, service access, weight, and repairability.

05Power: batteries, charging docks, tethering, hot swap packs, or vehicle power depending on the environment.

Real world applications

warehouse transport
hospital logistics
factory line supply
delivery
retail operations

Key technologies

autonomous navigation
fleet orchestration
safe obstacle avoidance
battery autonomy
dock charging

Sensors commonly used

2D LiDAR
3D LiDAR
RGB cameras
depth cameras
IMU
wheel encoders
safety bumpers
UWB beacons

Actuators or movement system

differential drive wheels
mecanum wheels
steered wheel modules
electric traction motors
braking systems
lift modules

AI and software used

SLAM
localization
path planning
obstacle avoidance
fleet management
battery management
remote diagnostics

Advantages

Can automate warehouse transport when the workflow is constrained and measurable.
Connects sensing, actuation, and AI into physical work.
Reduces exposure to repetitive, dirty, distant, or ergonomically difficult tasks.
Produces structured operational data that manual work rarely captures.
Improves when tools, fixtures, maps, and procedures are designed around the robot.

Current limitations

Performance drops when sensors face glare, dust, occlusion, deformable objects, poor lighting, water, smoke, or unexpected human behavior.
Hardware maintenance matters because motors, joints, seals, batteries, cables, and sensors degrade.
Most reliable autonomy is narrow and workflow specific.
Integration cost includes training, safety validation, spare parts, maps, network coverage, and support.
Human supervision is often needed for edge cases, recovery, cleaning, charging, or exceptions.

Popular examples and reference styles

tugger AGVs that pull carts on a milk-run route
unit-load AGVs with a roller or chain conveyor deck
under-rider or automated guided carts that lift and move a rack
pallet trucks that pick up and place floor-level pallets
automated forklifts for elevated pallet handling
custom heavy-load carriers for dies, coils or large assemblies

Deployment pattern

01Build a from-to matrix with load type, loaded and empty trip frequency, distance, peak demand and handoff time.

02Survey aisle width, intersections, floor joints, gradients, doors, elevators, charging locations and safe recovery zones.

03Simulate traffic and charging before buying vehicles; average demand can hide a peak-hour queue or deadlock.

04Pilot the complete loop with the intended load-transfer hardware and software integrations, not an empty vehicle demonstration.

05Run acceptance tests at rated payload and proposed speed, then track interventions and completed loaded missions for several representative weeks.

Evaluation metrics

01completed loaded missions per hour and per shift

02on-time pickup and delivery rate

03manual interventions per 100 missions

04blocked-route and safe-stop minutes

05queue time at intersections and handoff stations

06charge time and energy per completed mission

07mean time to recover and mean time to repair

08cost per completed loaded mission

Failure modes

01sensor occlusion or calibration drift

02unexpected object geometry

03battery or thermal limits

04network loss

05mechanical wear

06software edge cases

07operator confusion

Technical bottlenecks

01reliable perception in messy environments

02long duration autonomy

03safe contact with people and objects

04cost reduction without losing robustness

05data quality for robot learning

06integration with existing workflows

Research questions

01How can agv robots detect when their own perception is unreliable?

02Which tasks should be autonomous, teleoperated, or shared control?

03How can simulation produce behaviors that survive contact, lighting change, and hardware wear?

04What is the minimum sensor set that still provides safe and useful performance?

05How should usefulness be benchmarked instead of only showing impressive motion?

Safety, ethics, and responsible use

ISO 3691-4:2023 is the current published ISO safety standard for driverless industrial trucks and their systems; ISO lists a replacement draft under development. The standard covers AGVs and AMRs but does not make every installation safe by default. The operating zone, vehicle, attachment, load, traffic rules, protective devices, braking behavior, commissioning tests and foreseeable misuse all belong in the site-specific risk assessment. Never treat obstacle detection, a CE mark or an autonomy label as a substitute for validated stopping performance in the final facility.

Operator skills needed

basic robot safety and emergency stop behavior
understanding of maps, zones, missions, and task exceptions
daily inspection of sensors, batteries, cables, and end effectors
ability to read logs and distinguish robot failure from workflow failure
clear escalation process when autonomy is uncertain

Market signals to watch

buyers ask for uptime, service contracts, and measurable ROI
successful vendors simplify deployment and maintenance
robotics startups with data pipelines improve faster
large buyers care about safety, support, spare parts, and integration

Future potential

AGV systems are moving toward natural-feature navigation, richer diagnostics and mixed-vendor control. VDA 5050 version 3.0.0, released in March 2026, defines a current interface between mobile robots and central master control systems. Support still has to be verified by product and version; an interface claim does not remove the need for traffic engineering, safety validation or a recovery plan.

FAQ

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

A traditional AGV follows a defined guide path and normally stops when that path is blocked. An AMR typically localizes in a map and can plan or replan routes inside permitted zones. Vendor terminology overlaps, so compare actual navigation, obstacle, traffic and recovery behavior.

Does ISO 3691-4 certification make an AGV safe around people?

No single certificate replaces installation-level risk assessment and validation. Vehicle speed, payload, attachment, braking, protective fields, crossings, floor condition and operating-zone design must be tested together in the final application.

How should an AGV project calculate ROI?

Use the annualized vehicle, integration and infrastructure cost plus service, energy and recovery labor. Divide that amount by completed loaded missions. Compare it with the full cost and constraints of the existing material-flow process, and keep blocked or manually recovered missions visible.

Official sources and further reading

These primary and institutional sources support the technical descriptions in this guide. Product capabilities still vary by model, configuration and operating environment.

Recommended next guides

Robotics software Robot sensors Robot components

Related robot categories

Warehouse Robots / AMRs Delivery Robots Delivery Drones Educational Robots Research Robots Humanoid Robots Robot Sensors Robotics Topics

ROBOTICS RESEARCH

Go deeper intorobotics

Independent coverage of humanoid robots, Physical AI, industrial robotics, robot hardware and emerging automation systems.

Open the latest technical analysis, robot directories, and deployment guides built around measurable engineering decisions.

Explore Robotics News Open WaitAppoint

service@techniahqservice.com