A reality check: you don’t really “calibrate blind spots” in LiDAR like adjusting a software setting. In modern AGVs, blind spot handling involves:

• Safety zone configuration (software layer)

• Sensor parameter tuning (firmware layer)

• Environment correction (map + filtering)

• Sometimes physical sensor adjustment (hardware layer)

Most Chinese AGVs handle this through the RCS (Robot Control System) plus onboard safety controller.

Adjusting Safety Zones via Web Interface

Many mid-to-high-end Chinese AGVs allow partial adjustment of safety zones through:

• Web-based RCS dashboard

• Desktop configuration tools

• Engineering console (advanced users)

You can usually configure:

• Safety Zone Layers: warning, deceleration, emergency stop

• Radius / Shape: circular, polygon, or dynamic speed-dependent zones

Important: Critical safety parameters are often locked behind engineer/admin access due to ISO safety and liability requirements.

LiDAR Heat Maps

Advanced systems may show:

• Point cloud visualization

• Reflectivity or obstacle density maps

• SLAM confidence maps

These heat maps help engineers identify:

• Navigation instability zones

• False obstacle triggers

• Reflective interference areas

Lower-cost AGVs may only display basic maps and obstacle points without diagnostic heat maps.

Eliminating Ghost Objects

Causes of false LiDAR objects include:

• Dust or particles reflecting laser pulses

• Steam or condensation

• Temperature gradients

• Reflective shrink wrap or polished floors

Solutions (Layered Approach):

• Sensor-Level Filtering: noise thresholds, multi-frame validation, reflectivity filtering

• Software Filtering: temporal smoothing, obstacle persistence, minimum object size filtering

• Environmental Control: reduce steam, improve airflow, avoid direct exhaust

• Hardware Enhancements: LiDAR + vision fusion, dual LiDAR, radar-assisted filtering

Remote Support & Re-Scan

Chinese support teams can:

• Adjust SLAM map parameters and correct drift

• Resize safety envelopes and tune deceleration curves

• Connect via VPN/cloud to view real-time LiDAR streams

• Analyze point clouds remotely and suggest parameter adjustments

They cannot physically fix dirty lenses, misaligned sensors, or reflective structural issues remotely.

AGV System Architecture Overview

• Onboard Layer: LiDAR processing, emergency stop logic, local obstacle detection

• RCS Layer: fleet coordination, route planning, safety zone configuration, traffic control

• Cloud/Remote Layer (optional): diagnostics, logs, updates, remote tuning

Key Insight: Blind Spots

• Physical geometry: racks blocking LiDAR view angles

• Reflection environment: glass, metal, shrink wrap

• Software mapping: incomplete SLAM, incorrect safety zone tuning

Practical Deployment Tips

• Redesign aisle geometry slightly where possible

• Tune speed zones by area

• Segment warehouse into safety regions

• Combine LiDAR + vision validation

• Enforce strict mapping procedures during commissioning

What Buyers Are Really Asking

• Can my team maintain navigation accuracy independently?

• Will constant Chinese engineer involvement be required?

• Can the system adapt to dust, steam, and reflective surfaces?

• Is the software transparent for in-house management?

• How much is autonomous vs. vendor-dependent?

Bottom line: AGV hardware runs the robot, but software stability and mapping practices determine whether your warehouse operations run smoothly.