In 2026, the #1 reason AGV and AMR deployments run over budget or miss go-live dates isn’t the robot — it’s the sensor. A LiDAR that performs well in a demo but fails on a reflective floor, in a dusty aisle, or at full vehicle speed can turn a profitable project into a painful one.

This checklist is designed for systems integrators and robotics engineers who need to evaluate LiDAR readiness before committing to a deployment. Use it in your pre-project scoping, vendor evaluations, and site assessments.

Before anything else, validate that your LiDAR can reliably build and maintain a map of your facility — in the real environment, not just a demo space.

• Natural Feature Navigation — The LiDAR should scan existing facility contours (walls, racks, columns) to build and update maps — no tape, no reflectors, no infrastructure modifications required.

• 2D vs. 3D Selection — 2D LiDAR handles standard floor-level detection and SLAM. Choose 3D if your application involves creating digital twins, detecting overhead hazards, or multi-level environments.

• Dynamic Mapping — Verify real-time point cloud comparison against pre-loaded maps. The vehicle should maintain localization even as pallets, people, and equipment move around it.

• Environmental Validation — Test LiDAR performance under your specific conditions: high-glare flooring, low-light areas, dusty or humid environments, and reflective surfaces.

Safety failures don’t just injure people — they kill projects. This section is non-negotiable.

• Active Risk Reduction — The LiDAR must function as a primary safety sensor, not just a navigation aid. It must be capable of stopping the vehicle before a collision — not after.

• Programmable Safety Zones — Confirm support for configurable Warning zones (speed reduction trigger) and Protection zones (emergency stop trigger). These must be adjustable without hardware changes.

• ANSI/RIA R15.08 Compliance — Validate that both the LiDAR hardware and its integration meet the latest ANSI/RIA R15.08 standards for industrial mobile robots. Request documentation — don’t take verbal confirmation.

• Full-Width Coverage — The LiDAR’s detection zone must cover the full width of the vehicle including its maximum possible load. Test with your widest expected pallet or attachment.

Spec sheets look good on paper. Here’s what to actually verify before you commit.

• Range vs. Braking Distance — Match the LiDAR’s detection range to the vehicle’s maximum travel speed and calculated braking distance. At 2 m/s, you need more detection range than most data sheets will tell you upfront.

• Resolution for Your Application — High-resolution point clouds are essential for precision docking, navigating narrow aisles (< 1m clearance), and detecting low-profile obstacles like shrink wrap on the floor.

• Sensor Fusion Compatibility — Confirm the system supports data fusion with 3D cameras, IMUs, or UWB systems. Feature-poor environments — long corridors, large open bays — require sensor fusion to maintain reliable localization.

• Update Rate & Latency — Verify scan frequency (Hz) and data latency. Low-latency sensors are critical for vehicles operating at higher speeds or in dynamic environments with pedestrians.

A technically excellent LiDAR can still fail a deployment if the surrounding infrastructure isn’t ready. Validate these before your go-live date, not after.

• Network Capacity — Ensure your industrial Wi-Fi or 5G network can handle the sustained data stream from high-resolution LiDAR, especially in multi-robot fleet deployments. Bandwidth bottlenecks cause localization drift.

• Simulation First — Use LiDAR-generated data to run digital twin simulations of your logistics flows before go-live. This lets you optimize fleet size, routing, and zone layouts without expensive on-site testing.

• Mounting Position — Validate sensor mounting height for your use case: high on the chassis for navigation range; low for foot-level pedestrian detection. Many deployments require both.

• Integration Protocol — Confirm the LiDAR supports your industrial protocols: EtherNet/IP, PROFINET, or ROS2. Middleware translation adds latency and failure points.

3 LiDAR Mistakes That Derail AGV Deployments

Mistake #1: Speccing for the demo environment, not the real one. Sensor performance in a clean showroom has no bearing on performance in a facility with forklift exhaust, epoxy flooring glare, and seasonal temperature swings. Always request field test data from a comparable environment.

Mistake #2: Treating sensor cost as a line item, not a margin driver. A $1,500 difference in sensor cost across a 10-robot deployment is $15,000 off your margin. Over a five-year service contract, reliability differences compound significantly. Model the total cost, not just the purchase price.

Mistake #3: Skipping the safety zone validation. Many integrators configure safety zones in simulation but never walk the facility to verify real-world performance with actual floor surfaces and lighting. A missed pedestrian at low speed becomes a compliance investigation at high speed.

Quick Reference: Checklist Summary