Compact AMR in Narrow Aisles: How Body Size, Load Overhang, and Traffic Rules Determine Real Minimum Aisle Width

A spec sheet says “AMR width 600 mm.” A procurement team member checks that against an 800 mm aisle and figures there is plenty of room.

The robot clears the aisle — until it carries an 800 mm pallet. Or until it needs to turn 90° at the intersection. Or until a second robot needs to pass. Or until the safety LiDAR detects the wall and slows the robot to a crawl.

Robot width is just one input in the aisle width calculation. Real aisle width requirement = robot width + load overhang + motion margin + safety clearance + passing clearance (if bidirectional) + P&D action space. Every one of these variables can change the number significantly.

Adding Load Overhang and Carrier Geometry

Load overhang is the most commonly overlooked dimension in aisle width calculation.

The transport method determines overhang:

Illustrative example: Robot width 600 mm, carrying an 800 mm tote on top. Overhang per side = (800 − 600) / 2 = 100 mm. Maximum load width = 800 mm. One-way minimum aisle (with safety clearance) = 800 + 2 × clearance — the actual clearance depends on the robot’s safety configuration and your project requirements.

If the load is irregularly shaped (long workpieces, circular coils), overhang varies by position. Use the maximum overhang for calculation.

Turning, In-Place Rotation, and Lateral Movement

Aisle width needs to accommodate not just straight travel but also turning.

Different drive architectures have different turning space requirements:

Differential drive (two drive wheels + casters): The robot rotates around the drive wheel axis. The sweep path must be derived from the vehicle’s actual instantaneous rotation center and load envelope — request the supplier’s swept path diagram for the configured robot + load, and do not rely on a generic formula.

Omnidirectional drive (Mecanum wheels): Can move sideways. Turning does not require rotating the body, so the sweep area is smaller. But omnidirectional wheels change friction direction during lateral movement, demanding more from the floor. Loaded lateral speed may be lower than forward speed.

Four-wheel steering: Can reduce turning radius, but adds mechanical complexity and maintenance cost.

The key question for turning is not “can the robot make the turn?” but “does the maximum sweep area of robot + load during the turn exceed the aisle boundary?”

Ask the supplier for the swept path envelope diagram for the robot + standard load, then recalculate with your actual load geometry. If the supplier does not have this data, it needs to be measured during FAT.

One-Way vs. Bidirectional Traffic

Traffic rules directly determine width requirements.

One-way traffic: Robots travel in one direction only, no passing needed. Minimum aisle width = robot + load max width + safety clearance on both sides. (This is a planning worksheet formula; final clearance must follow the risk assessment, safety field configuration, and project geometry.)

Bidirectional traffic: Two robots need to pass each other in the aisle. Minimum aisle width = 2 × (robot + load max width) + center clearance. The actual multiplier depends on the robot dimensions, load geometry, and safety field configuration. (Planning worksheet formula; final clearance must follow the risk assessment, safety field configuration, and project geometry.)

Dynamic one-way: The physical aisle width does not support bidirectional traffic, but the dispatching system makes the aisle one-way in time — only one direction at any given moment. This is a common compromise in space-constrained warehouses, but it requires the dispatching system to support aisle-level direction locking and reduces throughput during peak times.

Which mode to choose depends on:

• Physical space constraints (can the aisle be widened?)
• Throughput requirements (does peak time need simultaneous bidirectional traffic?)
• Whether the dispatching system supports aisle-level direction control
• Fault tolerance (if a robot breaks down in a one-way aisle, can others get around it?)

P&D Actions Need More Space Than Travel

The space a robot needs to travel straight down an aisle and the space it needs to perform a pickup/drop-off (P&D) at the end of that aisle are two different numbers.

P&D actions typically include:

• Decelerating and positioning to the docking point
• Adjusting attitude (may require in-place rotation or lateral movement)
• Executing the pickup/drop-off action (lift, push, tow, etc.)
• Confirming completion (sensor detection)
• Accelerating away

The adjustment phase needs more space than straight travel because the robot’s sweep area during attitude adjustment may exceed the straight-travel envelope. If the docking point is at the aisle end, the robot may need to make a 90° turn before docking — requiring far more space than straight travel width.

Aisle width calculation cannot just consider “passing through.” It must also account for “what happens at the end of the aisle.” The additional space needed for P&D depends on the robot’s turning geometry, the docking approach angle, and the transfer method — calculate it for your specific robot and load.

Pedestrians, Blind Spots, and Safety Fields

When pedestrians share the aisle, width requirements increase further.

AMR safety LiDAR triggers slowdown or stop when it detects an obstacle. In a narrow aisle, the walls themselves are within the safety detection range. If the protective field is set incorrectly, the robot will continuously trigger slowdowns or be unable to move at all.

Common problems:

• Protective field wider than the aisle → robot constantly detects walls, persistent slowdown
• Blind corners → robot cannot see pedestrians or vehicles coming from the side during turns
• Pedestrian and robot aisles overlapping → safety risk and efficiency loss

The solution is not simply to shrink the protective field. Narrowing the safety field changes the robot’s stopping boundary — if the field is reduced, speed must also be reduced so that stopping distance fits within the new field. Any safety field adjustment must be:

1. Done by qualified personnel
2. Accompanied by a reassessment of speed and stopping distance
3. Verified on-site with actual load and traffic conditions

Dedicated pedestrian lanes, auxiliary sensors at blind corners, and speed limits in P&D areas are complementary measures — not substitutes for correct safety field configuration.

How to Measure the Real Narrowest Route

The “nominal” aisle width (what the drawing says: 800 mm) and the “effective” width (what is actually usable) are often different.

Effective aisle width can be eroded by:

• Shelf protrusions (beams, brackets, label holders)
• Floor marking offset
• Temporary stored items
• Columns and structural protrusions
• Cable trays and piping
• Door frames and thresholds

On-site measurement method:

1. Along the robot’s travel route, measure actual usable width at regular intervals.
2. Record all width constriction points (narrow sections) and mark locations.
3. Measure the actual usable area at turning zones (not the nominal area).
4. Confirm the effective space at P&D areas.
5. Mark all obstacles and their movability (permanent vs. temporary).

Narrow-Aisle Site Survey Checklist

Aisle Width Calculation Worksheet

Aisle Readiness Inputs

1. Do not compare only robot body width — aisle width requirement = robot + load overhang + motion margin + safety clearance.
2. Measure maximum load width, not robot width — the load may be much wider than the robot.
3. Turning sweep area is the key bottleneck — passing straight does not mean passing a turn.
4. P&D areas need more space than travel sections — calculate the additional space for your specific robot and docking approach.
5. Measure effective width on-site, do not rely on drawings — protrusions, temporary items, and marking offset erode nominal width.
6. Safety field configuration must match aisle width — narrowing the field requires reducing speed and reassessing stopping distance, not just shrinking the boundary.
7. One-way / bidirectional / dynamic one-way choice affects dispatching complexity — dynamic one-way is a compromise when space is tight, but requires dispatching system support.