AGV Traffic and Aisle Design

Uni-Directional vs. Bi-Directional Aisle Width

Standard AGV body width: 0.7-1.0m. Aisle width must accommodate body + sensor clearance, not just body.

Traffic Flow	Aisle Width	Notes
Single AGV, uni-directional	1.5m minimum	Standard nav systems
Two AGVs, bi-directional (passing)	2.5-3.0m	2x body + 0.5m clearance each side
High-density, single AGV	1.2m	Specialized nav + obstacle avoidance required

Passing clearance formula (bi-directional):

Bi-dir aisle = (vehicle width × 2) + (0.5m safety each side × 2)
Example: (1.0m × 2) + (0.5m × 2) = 3.0m

Bi-directional is more space-efficient in terms of travel distance (shorter routes) but requires 67-100% more aisle width than uni-directional. Most DC designs default to uni-directional flow to preserve storage density and simplify traffic control.

Traffic Control Methods

Zone Control (Most Common)

Aisle network divided into discrete non-overlapping zones. AGVs acquire a zone permit before entering. Only one AGV allowed per zone at a time.

• Simplest to install and expand
• Easy to debug and monitor
• Long aisles should be split into 2+ sub-zones so an AGV can clear an intersection while another waits mid-aisle

Forward Sensing Control

AGVs detect vehicles ahead using onboard sensors and stop or slow. No zone reservation required.

• Simpler infrastructure
• Higher risk of deadlock in bi-directional or multi-path networks
• More common in AMR systems than fixed-path AGVs

Combination Control

Zone control + forward sensing layered together. Most robust for dense fleets.

Deadlock Types and Prevention

Heading-On Deadlock

Two AGVs approaching from opposite directions on the same path. Neither can pass.

• Prevention: unidirectional flow design eliminates this entirely
• Mitigation: zone reservation logic rejects conflicting reservations before dispatch

Loop Deadlock

Three or more AGVs each trying to advance into the position held by the next, forming a circular block.

• Prevention: advance grid-block reservation (reserve path ahead, not just current position)
• Mitigation: three-layer traffic management (topological routing → path planning → real-time coordination)

Aisle Layout That Minimizes Deadlock

• Prefer unidirectional loop or race-track layouts over grid layouts
• Add dedicated passing bays at regular intervals on long uni-dir aisles
• Discretize long aisle sections into 2 sub-zones at midpoints
• Design crossing nodes with clear priority rules (one direction always yields)

Five Zone Types (ISO 3691-4 / RMI Framework)

Zone Type	Definition	Speed Limit
Operating zone	Normal travel, clearance >0.5m both sides	Up to 2 m/s
Operating hazard zone	Active scanner present, reduced clearance	Up to 1.2 m/s
Restricted zone	Clearance <0.5m both sides, fixed racking	Up to 0.3 m/s
Confined zone	Enclosed space, special access controls	Site-specific
Load transfer area	Pick/deposit stations	Creep speed only

Mixed Traffic: AGV + Manned Forklift

Do not share rack aisles without interlocks. A manned forklift in an AGV rack aisle requires:

• Traffic light / zone interlock at aisle entry: AGV pauses when manned truck enters
• Physical gate or barrier preferred over software-only control

Cross-traffic zones (transfer aisles, receiving docks): design with AGV crossing points separated from manned travel lanes by minimum 1.5m marked pedestrian/equipment corridor.

Mixed Traffic: AGV + Pedestrian

Layered controls in order of effectiveness (most to least):

1. Physical barriers (guardrails, bollards, gates): most effective; eliminate contact risk
2. Floor marking + striping: OSHA baseline; inadequate alone for AGV environments
3. Speed/right-of-way rules: AGV always yields to pedestrian detection
4. Visibility aids: LED warning lights on AGVs (not standard — specify at procurement)
5. Detection alerts: audible/visual alarms at zone entries

Pedestrian crossing points should be perpendicular to AGV travel direction, not parallel. Install light curtains or safety mats at crossings that pause AGV when tripped.

Traffic Capacity Model

Each aisle has a finite traffic capacity measured in AGV-moves/hour. Each vehicle consumes capacity proportional to its run time plus dwell time at load/unload stations:

Capacity consumed per AGV-trip = travel time in aisle + dwell time at station
Aisle utilization = (fleet trips/hr × avg consumed capacity) / total aisle capacity

To maximize system throughput, formulate as a constrained optimization:

• Decision variables: number of AGVs, route assignments
• Constraint: each aisle’s traffic capacity is not exceeded
• Objective: maximize total moves per hour

Practical implication: adding more AGVs beyond the aisle capacity constraint does not increase throughput — it increases congestion and dwell. The binding constraint is almost always the aisle with the highest demand, not the fleet size. Source: IEEE Xplore (Ieeexplore.ieee.org/document/7889626) — medium confidence (abstract only reviewed).

Throughput Design Rules

• Design around bottlenecks, not average aisle capacity. The narrowest point in the network limits fleet throughput regardless of total aisle capacity.
• Size fleet to peak-hour demand + variability, not daily average moves. Peak:average ratios in DCs often run 2:1 to 3:1.
• Throughput gains come from control boundary efficiency, not adding vehicles. Tighten PLC handshakes, reduce transfer dwell, improve route reservation logic before expanding fleet.
• VNA conversion: switching from reach-truck aisles to VNA aisles in the same building can increase storage capacity by up to 50%.
• Reserve aisle width for growth: design to current vehicle + one size class up. Widening aisles after racking is installed means re-slotting.