Warehouse Monitoring Solution Introduction

type

status

date

slug

summary

category

icon

password

Challenge: In most of the warehouses where AGVs have been applied, there are still cases of manual stacking of goods, which will lead to the WMS system not being able to determine the actual occupation of each storage space in real-time, which will result in the inaccuracy of the information provided by the WMS to the AGVs. If the occupied storage space cannot be recognized in time, the AGV may receive wrong commands, which not only reduces the operation efficiency, but also easily collides with other goods or equipment, and in serious cases, it may even lead to safety accidents, endangering the safety of personnel and equipment.

The Warehouse Monitoring system for storage positions has emerged as a result. The system optimizes the allocation of storage resources, reduces human errors, and improves operational efficiency through functions such as automatic positioning, real-time data synchronization, and precise inventory control.

Pain points of two common types of location status recognition sensors:

The location information transmitted by the single-point laser radar is not accurate enough, which can easily cause stacking accidents

Some warehouses use single-point lidar for location detection, but this ranging radar can only emit one laser beam at a time to form a point on the surface of the object, which may ignore the gaps between cartons or pallets, leading to misidentification of the location status and causing stacking accidents.

The verification method of pure RGB cameras is single, and there is a risk of misjudgment and information loss

There are the following drawbacks when using RGB cameras to determine the status of storage locations:

When deep learning detects targets, objects outside the training dataset entering the database may cause false detection and provide incorrect information.

The lack of cargo height information makes it difficult to arrange stacking tasks.

Ultra-wide-angle fisheye cameras have edge distortion problems in location determination, which affect the accuracy of model training and prediction, and increase server costs.

Solution Overview

MRDVS uses RGB-D cameras to provide three-dimensional data and color information on warehouse locations, automatically identifying the presence, placement specifications, and abnormal occupancy of goods. It can also distinguish goods categories through the 3D vision intelligent AI system, output to the scheduling system, and achieve intelligent in-and-out inspection. The camera has built-in computing power and does not require an external industrial computer. At the same time, MRDVS provides a software and hardware integrated solution for warehouse location status recognition, as well as professional service support such as model training and deployment guidance.

Core Advantages

Comprehensive monitoring: Through the collection of three-dimensional data and color information, combined with AI technology, accurately identify the status of the warehouse.

Easy deployment: The database status recognition algorithm is placed on the camera side to reduce deployment and maintenance costs.

Flexible communication: Supports multiple communication methods such as TCP/IP, UDP, HTTP, etc., and reports data to the control system in real-time in JSON format.

Improve efficiency: Real-time transmission of location information to assist the scheduling system in quickly and accurately assigning tasks.

Animated Demo

MRDVS Warehouse Monitoring Industry Solution

Camera selection advice

S2 Max

Camera installation method

Fixed installation: It is recommended to install the camera directly above the monitoring location when it is installed on a pillar or ceiling.

Specific workflow

Step 1: The upper-level control system sends task instructions, and the 3D vision camera captures images and accurately obtains the specified position status.

Step 2: The detection results of the storage location are reported to the upper system through JSON to assist the scheduling system in achieving intelligent storage management.

Communication methods: TCP/IP, UDP, HTTP