Seven navigation families now cover the AGV market, with magnetic-tape, laser, vision, QR-code, inertial, inductive-wire, and SLAM-based units each tied to specific payload, accuracy, and integration limits [S1][S3].
Fleet-level coordination is converging on the VDA 5050 message set, supported by 100+ AGV/AMR builders using a common automation platform with vehicle controllers, navigation software, and a system manager for mixed-vendor fleets [S3].
Navigation Technologies Compared: Accuracy, Floor Prep, and Retrofit Cost
AGV navigation runs across seven physical principles: buried inductive wire, surface-applied magnetic tape, reflective laser targets, floor QR/2D-code grids, inertial measurement with map correction, on-board vision landmarks, and laser SLAM without fixed reflectors [S1][S5]. Laser-guided AGVs and vision-guided vehicles dominate modern greenfield builds, while magnetic-tape and QR-code units remain common in brownfield retrofits where floor modification is constrained [S1].
Specifiers should weigh three numbers: positioning repeatability (typically ±10 mm for laser, ±25 mm for tape, ±5 mm for QR/SLAM in structured light), floor-prep cost (tape ≈ $2–5/m, inductive wire ≈ $15–30/m, QR tags ≈ $0.20–0.50 each), and re-routing flexibility (SLAM = highest, inductive wire = lowest) [S1][S3]. Kollmorgen's NDC platform cites 50+ years of AGV fleet-controller development, with current offerings covering CVC700 vehicle controllers, NDC Flex navigation, and System Manager NG for VDA 5050 fleet dispatch [S3].
Vehicle Body Types and Their Working Payload Bands
Five AGV body types map to five payload classes: tow tractors (up to 6 t, train multiple carts), unit-load forklifts (1–2.5 t, pallet handling), low-profile underride AGVs (200–1 500 kg, conveyor or roller deck), assembly-line mobile racks (50–500 kg, automotive trim lines), and small bin-handlers (10–100 kg, e-commerce piece picking) [S1][S4]. Lisen Automation's product line explicitly lists laser-guided AGVs, vision-guided AGVs, and tow-type vehicles as its three standard body forms, supporting storage, picking, and inter-process material flow [S1].
Forklift-style AGVs dominate pallet moves because they integrate forks, lift masts, and battery packs in one chassis; tow tractors dominate long-run milk-runs where one tractor pulls 3–10 trailers to fixed pickup/drop stations [S1]. For an overview of how mobile robots fit plant-wide, see the AGV robot reference page and the industrial valve reference for fixed plant equipment that AGV fleets typically tie into via MES handshakes.
Application Segments: Where Each AGV Class Pays Back
Automotive manufacturing, e-commerce warehousing, pharmaceutical cleanroom logistics, food and beverage palletising, postal/parcel sorting, and battery/EV plant flow are the six highest-density AGV application segments in 2026 [S3][S4]. Lisen Automation's published use cases include production-line inter-process moves, warehouse storage optimisation, and distribution-centre picking under configurable guidance paths [S1].
Cleanroom and pharmaceutical AGVs typically demand stainless or ESD-safe bodies, IP54–IP65 sealing, and cleanroom-class particulate counts; pallet-handling AGVs in food plants need washdown-rated frames and food-grade lubricants [S3]. For related spec work, the pressure transmitter reference covers the fixed instrumentation that often sits on the same skid as an AGV charging station.
Selection Criteria: Payload, Aisle Width, Floor Condition, and Fleet Scale
Four criteria decide the AGV class: rated payload versus peak load, minimum aisle width versus vehicle turning radius, floor condition versus allowable markings, and fleet size versus the need for a central traffic manager [S1][S3]. AGVs with on-board vision localisation work best in dynamic warehouses where the floor cannot be taped, while inductive-wire units still win in clean, unchanging corridors where reliability outweighs flexibility [S1][S5].
Fleet scale matters because a 5-vehicle fleet can run on a single controller, but anything above ~20 vehicles needs a fleet manager handling intersection precedence, battery-state routing, and VDA 5050 message exchange with WMS/MES [S3]. Magnet-and-tape AGVs hit the lowest unit price, but the reconfiguration cost grows linearly with route metres, which is why SLAM-based units are winning new greenfield bids even with a higher upfront price [S1].
Standards, Fleet Protocols, and Safety Guarding
AGV safety falls under ISO 3691-4 (driverless industrial trucks), with EN 1525 for older installations and regional marks such as CE for Europe; VDA 5050 is the de facto fleet-manager-to-vehicle message protocol, exposing order, state, and visualisation messages to a mixed-vendor fleet [S3]. Functional safety is typically built around PL d / SIL 2 laser scanners, bumpers, and e-stops per the relevant ISO clauses, with safety-rated PLCs interfacing to the vehicle controller [S3].
Battery chemistries split between lead-acid (lowest capex, 6–8 h cycles), lithium-iron-phosphate (LiFePO4, 3 000+ cycles, opportunity charging), and increasingly sodium-ion for cold-store work where low-temperature performance matters [S1][S4]. For fixed-instrumentation work that AGV controllers often query via OPC UA, see the flow meter reference and the pressure sensor reference.
Limits, Failure Modes, and 2026 Market Signals
AGVs are not a fit for very low-volume SKUs, highly variable tote sizes, or outdoor unpaved yards, where AMRs with SLAM or even manned forkloves still win on flexibility and unit cost [S1][S3]. The hard limit on AGV payback is route repeatability: an AGV that runs the same loop fewer than 8–10 times per shift rarely justifies the navigation infrastructure, regardless of vehicle class [S1].
Two signals to watch through late 2026: VDA 5050 2.0 message-set adoption by European car OEMs, and the move by Chinese tier-1 integrators (Lisen Automation among them) to ship vision-guided AGVs as a stocked item rather than a custom build [S1][S3]. For plants already spec'ing mobile fleets, the electric ball valve selection map covers the fixed-valve side of the same automated line.