What is an AGV — the plain-language definition
An Automated Guided Vehicle (AGV) is a battery-powered, driverless vehicle that transports materials through a warehouse, factory, or distribution centre along a defined route — automatically, without a human operator, and without deviating from its programmed path.
The simplest way to understand an AGV is to think of a train on a track. The track defines exactly where the train can go. The train does not decide its route — the route was decided when the track was laid. An AGV works the same way: the route is defined in advance (through magnetic tape, laser reflectors, or embedded wires), and the AGV follows that route repeatedly and reliably.
AGVs have been in industrial use since Barrett Electronics introduced the first one in the 1950s, guided by embedded wires in automotive plant floors. Today, the global AGV market is valued at USD 5.93 billion in 2025, growing to USD 11.58 billion by 2033 (Grand View Research). They are deployed in automotive manufacturing, pharmaceutical warehouses, food and beverage production, cold storage, e-commerce fulfillment, and anywhere high-volume, repetitive material transport needs to be automated reliably.
How an AGV works — step by step
Understanding how an AGV operates requires understanding four separate systems working together: the navigation system (how it knows where it is), the fleet management system (how it receives tasks), the movement and steering system (how it physically moves), and the safety system (how it avoids accidents). Here is how all four work in sequence.
The 5 AGV navigation technologies — explained precisely
Navigation is the defining technology of an AGV — it determines installation cost, flexibility, accuracy, and what happens when the environment changes. There are five navigation methods in commercial use, ranging from simple floor tape to sophisticated SLAM-adjacent natural feature recognition.
Magnetic tape navigation
How it works: Magnetic strips (tape) are adhered to or embedded in the floor surface, creating a physical guidepath. The AGV's underside sensors detect the magnetic field continuously and steer to stay centred over the strip.
Speed: Typically 1–2 m/s on magnetic tape. Infrastructure: Low — tape can be surface-mounted without floor modification. Embedded wire systems require cutting floor channels. Accuracy: High along the defined path (±5–10mm).
Rerouting: Expensive and disruptive — requires laying new tape and reprogramming. Any layout change means physical tape modification. Best for: Simple, stable routes in small to medium facilities; first implementation, low capital. Cost profile: Lowest installation cost of all navigation types.
Laser guidance (LiDAR + reflectors)
How it works: Retro-reflective targets are mounted at known positions on walls, pillars, or racking. The AGV emits a rotating laser beam that reflects off these targets. By measuring the angle and return time of multiple reflections simultaneously, the AGV calculates its exact position through triangulation — typically accurate to ±5mm.
Infrastructure: No floor modification required — only wall-mounted reflectors at known coordinates. Flexibility: Higher than magnetic tape — routes can be reprogrammed in software without physical changes (reflector positions remain fixed, route logic is updated). Speed: Up to 3–4 m/s.
Market position: Laser guidance holds the largest market share in new AGV installations globally at 39%+ (Grand View Research). Best for: Medium to large facilities, precise positioning requirements, environments where floor modification is not possible.
Optical / vision navigation
How it works: Downward-facing cameras read visual markers — painted lines on the floor, QR codes at intersections, or coloured guides. The AGV's image processing software interprets the visual input and steers accordingly. Some advanced systems use forward-facing cameras to recognise broader environmental features.
Infrastructure: Painted lines are low-cost but require the floor to be clean and lines to be maintained. QR codes at intersections provide more reliable navigation. Limitation: Performance degrades if lines are worn, dirty, or obscured. Lighting conditions must be consistent. Best for: Environments where floor marking is practical and maintained — some food/beverage and pharma applications.
Inertial guidance
How it works: Gyroscopes and wheel encoders track the AGV's movement from a known starting position — measuring every turn, acceleration, and distance travelled. This is called dead-reckoning. Because small errors accumulate over time, inertial AGVs use floor-embedded magnets or QR codes at defined intervals to correct their position estimate periodically.
Accuracy: Good in the short term, requires regular correction points for longer routes. Infrastructure: Minimal floor modification — only periodic correction markers. Best for: Medium-distance routes with periodic correction point opportunities. Often combined with magnetic markers.
Natural feature navigation (no markers)
How it works: The most advanced AGV navigation method. The AGV uses onboard sensors to build a map of its environment by recognising existing features — walls, columns, racking uprights — without any installed markers. Position is calculated by continuously comparing sensor readings to the stored map. This method is essentially the same as the SLAM (Simultaneous Localization and Mapping) technology used in AMRs — the distinction between an advanced natural-feature AGV and an AMR is increasingly one of flexibility (whether the vehicle can dynamically re-route) rather than navigation technology alone.
Infrastructure: Zero — no floor modification, no reflectors, no markers. Flexibility: High — environment can change and the AGV can be retaught its route by walking it through once. Best for: Dynamic environments, facilities where routes change frequently, or where no physical installation is permitted.
The 6 types of AGVs — and what each is used for
AGVs are not a single product — they are a category. Within that category are six distinct vehicle types, each designed for different load types, weight classes, and operational environments.
Type 01 — Tow vehicle (tugger)
Heavy loads · Long distancePulls multiple cargo-bearing trailers linked together in a train formation. A single tow AGV can pull 3–10 trailers, transporting large volumes of material in one pass. Largest AGV segment globally at 38%+ of market revenue.
Type 02 — Unit load carrier
Pallets · Containers · TotesCarries a single discrete load on a flat platform. Often equipped with a lifting mechanism (hydraulic or scissor lift) to pick loads from the floor and deposit them at defined locations. Common in warehouse transport between storage and production areas.
Type 03 — Pallet truck AGV
Standard palletsAutomated version of a manual pallet jack. Forks slide under a standard pallet at floor level, lift, and transport to destination. Cannot access elevated rack storage — operates only at floor level. Simple, cost-effective for floor-to-floor pallet movement.
Type 04 — Assembly line vehicle
WIP · ComponentsTravels slowly alongside production lines, carrying work-in-progress items or component kits from station to station. Speed matched to the assembly line rhythm. Keeps assembly workers supplied without manual material handling. The original use case for AGVs — automotive plants in the 1950s.
Type 05 — Forklift AGV
Rack storage · High liftAutomated forklift capable of lifting pallets to rack height for storage and retrieval in multi-level racking. The most complex and expensive AGV type — requires precise positioning to engage pallet slots at height. Increasingly combined with laser navigation for sub-centimetre accuracy at rack level.
Type 06 — Automated Guided Cart (AGC)
Small loads · SortingThe simplest AGV type — small, lightweight platforms designed for light loads in controlled environments. Often used in sorting operations, pharmaceutical distribution, and electronics component transport where loads are small but transport frequency is high.
Evaluating warehouse automation for your operation?
Fast WMS integrates with AGV and AMR fleet management systems via standard API — the WMS generates tasks, the robots execute them. A demo shows exactly what the WMS layer provides.
How AGVs integrate with a WMS
An AGV does not operate in isolation. It is the physical execution layer of a system whose brain is the WMS. Understanding how they connect is essential for anyone evaluating AGV deployment.
The integration follows a two-layer architecture:
WMS — Warehouse Management System
Generates transport tasks from business events: a sales order triggers a pick task, a GRN triggers a put-away task, a production work order triggers a material supply task. The WMS defines what needs to move and where it needs to go.
FMS — Fleet Management System
Receives tasks from WMS. Assigns each task to the optimal AGV based on current vehicle position, battery level, and queue depth. Manages traffic control to prevent collisions between multiple AGVs. Manages charging schedules — AGVs returning to charge stations when battery drops below threshold. The FMS defines who does the task and when.
AGV — Physical execution
Receives movement commands from FMS. Executes: navigates to pickup, handles load, transports to destination, deposits load, sends completion signal back to FMS → FMS updates WMS → task marked complete, stock movement recorded.
The emerging standard for WMS-to-AGV/AMR integration is VDA 5050 — a vendor-neutral interface specification that allows AGVs and AMRs from different manufacturers to be managed through a single Fleet Management System. For Indian businesses evaluating AGV deployments, VDA 5050 compliance is worth confirming with vendors — it prevents lock-in to a single AGV manufacturer's proprietary interface.
AGV vs AMR — the honest comparison
The most common question after understanding AGVs is: should we choose an AGV or an AMR? The honest answer is not "one is better" — it's "they are different, and each is better in specific conditions."
The comparison above shows differences, not winner and loser. In a stable, high-volume automotive assembly plant where the same pallet moves between the same two points 300 times a day, an AGV's predictability and heavy-load capability are advantages. In a 3PL warehouse where layout changes quarterly and human workers share the floor, an AMR's flexibility and obstacle avoidance are more important.
An increasingly common approach is the mixed fleet: AGVs for fixed heavy-load routes and AMRs for dynamic picking and flexible transport, both coordinated through a single fleet management system with VDA 5050.
When to choose AGV over AMR (and vice versa)
A practical decision framework — based on operational characteristics, not technology preference.
| Decision factor | Choose AGV | Choose AMR |
|---|---|---|
| Layout stability | Fixed, won't change for 3–5+ years | Changes frequently or seasonally |
| Task type | Same route, same load, high repetition | Variable routes, variable loads, multiple task types |
| Load weight | Heavy (>1 tonne) or very heavy (multi-tonne) | Light to medium (most AMRs under 1 tonne, exceptions exist) |
| Human coexistence | Separated areas — AGVs on fixed lanes | Shared floor with workers — AMR safety systems designed for this |
| Deployment urgency | Weeks–months (infrastructure install) | Days–weeks (no floor modification) |
| Environment | Controlled, predictable, no obstacles | Dynamic, changing, people and equipment sharing the space |
| Scalability | Add AGVs = add fixed routes (infrastructure) | Add AMRs = add units + update software (no infrastructure) |
| Budget type | Higher CapEx upfront (infrastructure), lower per-unit | Lower initial CapEx, higher per-unit cost |
| India context best fit | Automotive assembly, large FMCG distribution, pharma (GMP) | 3PL multi-client, e-commerce fulfillment, mixed-use warehouses |
AGVs in India — which industries, what's deployed
India's AGV market was valued at USD 174.4 million in 2025, with IMARC projecting growth to USD 464.4 million by 2034 at 11.15% CAGR. The Asia Pacific region, including India, is the fastest-growing AGV market globally. Here is what is actually deployed in India by sector.
India automotive sector received USD 36.268 billion FDI (April 2000–March 2024). Tata, Mahindra, and Hyundai India (Chennai) are integrating AGVs into assembly lines — tow vehicles supplying component kits, assembly line vehicles pacing with production speed. EV production push (government target: 30% EV sales by 2030) is driving AGV integration into battery assembly and component logistics. Automotive accounts for the largest share of India's industrial robot and AGV demand.
Schedule M compliance (non-negotiable from January 2026) requires GDP-compliant material handling with minimal human contamination risk. Forklift AGVs and unit load carriers are being adopted for moving product in controlled temperature zones — eliminating human operators from GMP-critical areas. Full chain-of-custody documentation (required for audit) is generated by WMS integration with the AGV task record.
Flipkart's Bengaluru delivery hub: AGV-based sorting robots (GreyOrange) process 4,500 packages per hour — ten times manual throughput. This is the most visible AGV deployment in India. Other large e-commerce operators (Amazon India, Delhivery sortation hubs) are expanding robotic sorting infrastructure. This segment is the fastest-growing for AGV deployment in India.
Temperature-controlled facilities restrict human access to minimise thermal fluctuation. AGC-type robots and unit load carriers operate in chill zones to move goods between storage and packing without human entry. FEFO enforcement is managed by WMS, task assignments sent to AGV fleet. Solar-integrated cold storage (India guidelines 2025) is adding power-efficient AGV charging infrastructure.