A Method of Center of Gravity Assessment Based on Object Scanning and a Method of Grabbing Object with a Robot There Of

HKSAR

32024092278.3

Robotic Perception and Positioning System for Sorting Precision Fine-grained Object

HKSAR / PRC

HKSAR: 32024096292.0 / PRC: 2024112757393

Robotic perception and positioning system for sorting precision fine-grained object

CargoFlow: Advanced Congestion Detection and Root Cause Analysis System for Air Cargo Logistics

US

63/689,775

Reflective Device For Surface Defects Detection

HKSAR

32023083911.2

The present invention relates to a system for capturing defects on a surface of an object in a single-camera setup. By using multiple cameras, different perspectives of the object can be obtained, allowing for better coverage and a more comprehensive examination of the surface. As irregular shapes may have varying depths, concavities, or protrusions that cannot be adequately captured by a common single camera setup, multiple cameras are therefore positioned at different angles to enable better depth perception and accurate identification of defects in three-dimensional space.

一種基於一維resunet的電梯振動信號去噪方法

PRC

2410021289.9

本申請涉及電梯技術領域，特別涉及一種電梯振動信號去噪方法、裝置、電子設備及存儲介質，其中，方法包括：採集電梯的轎廂垂直振動信號，並生成原始一維時間序列信號; 將不同類型、不同強度的雜訊注入到原始一維時間序列信號，得不同類型的一維含噪混合信號; 採用滑動視窗方式分割生成目標長度的數據樣本並劃分為訓練集、驗證集和測試集; 構建ResUNet深度神經網路並採用訓練集對ResUNet訓練直至網路收斂，採用驗證集對超參數調整，獲得訓練完成的網路模型; 將測試集的含噪振動信號輸入到訓練完成的網路模型進行預測，輸出去噪后的振動信號。 由此，解決了相關技術中，由於需要依賴電梯振動信號的先驗知識或參數的人為設定，影響去噪的有效性和可靠性等問題。

Mechanical Gripper And Method For Using It To Pick And Place Objects

HKSAR / PRC

HKSAR: 32023075720.7 / PRC: CN 202310853939.1

The present invention discloses a mechanical gripper. The mechanical gripper comprising a mounting bracket and a plurality of suction parts arranged in an array on the mounting bracket for suctioning objects, wherein each suction part has an adjustable length along its suction direction. The mechanical gripper provided by the present invention has good versatility. The present invention also discloses a method of using the mechanical gripper to pick and place objects.

自動引導車路徑規劃方法、電子設備及存儲介質

PRC

202410439782.2

本申請實施例提供了一種自動引導車路徑規劃方法、電子設備及存儲介質，屬於路徑規劃技術領域。 該方法包括：從網格網路中篩選出初始網格的一跳網格，得到第一候選網格。 構建第一候選網格的第一搜索狀態資訊，並保存到佇列中。 獲取佇列中第一最早時間戳最小的第一搜索狀態信息，對應的第一候選網格為第一目標網格。 若第一目標網格不屬於預設的目標位置，和/或當前運動狀態為移動狀態，則篩選出第一目標網格的一跳網格，得到更新的第一候選網格。 重複上述步驟，直至第一目標網格為目標位置，且當前運動狀態為靜止狀態。 根據初始網格的第一搜索狀態資訊和所有選擇出的第一搜索狀態資訊構建自動引導車的運動路徑，提高了AGV的實際運行效率。

A Method For Realizing Autonomous Navigation Of Robots Using Bim

HKSAR

32023079478.8

A Method for realizing autonomous navigation of robots using BIM, first, the BIM model is used to generate a navigation map, and then the site is optically scanned through 3D scanning to obtain on-site point cloud images and semantic segmentation is performed to obtain a semantic map. The robot uses the semantic map to achieve autonomous navigation. The method of using BIM model to map conversion and semantic understanding to realize autonomous navigation of robots can overcome the obstacles of existing robot platforms in realizing automation and other related fields, and increase the autonomy of robots.

Method for Detecting and Predicting a Bottleneck in a Transportation Process of a Logistics Centre

HKSAR

32023070062.9

System And Method For Detecting Surface Defect Of Object

HKSAR / PRC

HKSAR: 32023070552.9 / PRC: CN 202310309532.2

Defect Detection Method And Device Based On The Training Of The Dataset With Few Normal Samples

HKSAR / PRC

HKSAR : 32022066158.3 / PRC: CN202211689395.1

A 3D-Reconstructed Object Surface Inspection Method Based On A Robotic Arm And Eye System

HKSAR / PRC

HKSAR : 32022062462.3 / PRC: CN202211288329.3

Disturbance Detection And Countermeasure Suggestion System And The Method Therefor

HKSAR

32023083910.4

The present invention relates to a system (100) and a method (200) to detect and classify disturbances and derive countermeasures, characterised by: a data component module (101), a deviation reporting module (1021) consisting of a multi-level data analysis module (1021), a deviation detection module (1022) and a conformance checking module (1023), a disturbance diagnosis module (103) consisting of a classifier module (1031) and a countermeasures module (1032); and a user interface module (104) configured to display disturbances obtained from the countermeasures module (1032), wherein disturbances are detected, diagnosed and classified using the deviation reporting module (102) and the disturbance diagnosis module (103) for automatically suggesting countermeasures.

A Quality Prediction And Root Cause Analysis System For Injection Molding Processes

HKSAR / PRC

HKSAR: 32022062264.3 / PRC: CN202310827739.9

The present invention relates to a quality prediction and root cause analysis system. In particular, the invention pertains to an AI-based system for analysing various data obtained in previously executed manufacturing operations for performing quality prediction and root cause analysis to determine one or more possible root causes for production deviation resulted in a manufacturing process.

A Multi-Sensor Tool Wear Condition Prediction And Monitoring System

HKSAR / PRC

HKSAR: 32023071003.2 / PRC: 202420628565.3

The present invention discloses a multi-sensor (102) tool wear condition prediction and monitoring system, comprising of: a plurality of sensors (102) for processing analogue signals of the sensors (102); wherein the plurality of sensors (102) comprises an acoustic emission sensor (102a), a vibration sensor (102b) and a current sensor (102c); an IOT enabled remotely accessible user interface (118) with machine (118a) and scheduling frontends (118b); a software hub (202), characterized by: a data acquisition module (108) connected to the plurality of sensors (102), a data pre-processing module (114), a data storage module (202a) and a device management module (202b); and a machine learning prediction model (116).

System And Method For Cnc Tool Wear Conditions Monitoring And Prediction By Data Augmentation

HKSAR

32023067677.9

The present invention discloses a tool wear condition prediction and monitoring system based on machine learning, comprising of: a user interface (102) with machine (102a) and scheduling frontends (102b); a software hub (104) for providing data collection (104a), data processing (104b), data storage (104c) and device management (104d); a machine learning prediction model (106), characterized by; a feature extraction module (106a) configured to extract features; a data generation module (404) configured to generate synthetic features (302); and a prediction module (406) that is configured to predict a tool wear condition.

Method And System Of Cloud-Based Cad Automatic Assembly Sequence Planning

HKSAR

32023070184.1

The present invention discloses a system of automatic assembly sequence planning for cloud-based computer-aided design (CAD), comprising of: a computer device of a CAD model having a processor and non-transient media; a cloud-based computer server having a processor and non-transient media for a cloud-based CAD software program for assembly sequence planning (102); wherein the cloud-based CAD software program for assembly sequence planning (102) includes instructions for: data extraction (302); automatic constraint modelling; optimized assembly sequence generation; visualization of optimal assembly sequence (312); and a user interface for providing an optimal users interaction in the cloud-based CAD software program for assembly sequence planning (102).

Failure Risk Prediction System for Platform Screen Doors of Train

HKSAR / PRC

HKSAR : 32023068412.0 / PRC: 202310165356.X

The present application discloses a failure risk prediction system for platform screen doors of a train, which includes: a plurality of sensors arranged at predetermined positions of the screen doors for collecting real-time data; a gateway located near the sensors to transmit the data collected by the sensors to an edge device which is used to process the data transmitted from the gateway and transmit the processed data to a server, a deep learning model is used to predict the failure risk of the screen doors according to the log data of the screen door; the server, used to store data and train the deep learning model; the environmental data of the screen door, and the operation data of the screen door, and the operation data of the screen door is collected by the sensors. The present application can be used for predictive maintenance of platform screen doors of the train.

Method and System of Automatic Constraint Modelling for Assembly Sequence Planning

HKSAR

32023067195.2

The present invention discloses a comprehensive assembly sequence planning based on CAD data and a fully automatic pipeline to achieve a reliable assembly sequence. An automatic constraint modelling (100) method for assembly sequence planning is disclosed in the present invention wherein, comprising the steps of: recognizing joining parts of a product data using a feature recognition (102) algorithm to automatically recognize shape and determine assembly features; dividing the product data into subassemblies (104); deriving liaison matrix for parts inside each of the subassemblies (104); deriving interference matrices by analyzing potential collisions (108) of the parts for each of the subassemblies (104); and defining and selecting base component (110) of the subassemblies (104). After the automatic modelling (100), the constraints are introduced to the Optimal Assembly Sequence Generation (112) where assembly sequences are generated based on the constraints and evaluated (114) regarding the optimization criteria.

Device for the Optical Inspection of Ultrasonic Welds

PRC

202222999357.8

The present invention relates to a device for the optical inspection of ultrasonic welds, particularly for the inspection of welds about the periphery of thermoplastic parts forming an envelope.