3D Random Bin Picking Technology for Industrial Robot (ART/220CP)

3D Random Bin Picking Technology for Industrial Robot (ART/220CP)

  • 3D Random Bin Picking Technology for Industrial Robot (ART/220CP)
    ART/220CP
    Platform
    01 / 12 / 2016 - 31 / 07 / 2018
    9,600

    Ms Anna Ying LIU

    1. Develop 3D data acquisition systems, including (1) Design coded phase shift projection patterns for purpose of high-performance 3D data acquisition systems in terms of speed, accuracy and measurement range (2) Develop optical engine of 3D data acquisition system excluding camera (contrast: > 400; Image quality: MTF > 0.2 @ 37lps/mm; F number: 3.5) (3) Develop 3D point cloud generation algorithms based on coded phase shift projection patterns (4). 3D data acquisition systems integration Based on the above three technologies, 2 types of 3D data acquisition systems will be delivered for different applications. (a) Narrow-field 3D data acquisition system Field of view (FOV): 100 ~ 340 mm Accuracy: < 0.2mm @1m Speed: 0.5s Measurement range: <1 m (b) Wide-field 3D data acquisition system Field of view (FOV): 248 ~ 828 mm Accuracy: < 0.5mm @1m Speed: 0.5s Measurement range: <1 m 2. Develop 3D object recognition technologies, including (1) Develop fast feature extraction method from 3D data generated by 3D data acquisition systems (2) Develop machine learning based method to identify parts, measure part location and estimate part posture based on the extracted features Based on the above two technologies, 3D object recognition algorithms will be delivered with the following specifications. Recognition time: < 1s Location repeatability: +/- 0.1mm 3. CS deliverables for Customer (2) 1 set prototype of 3D data acquisition system with features: Wide FOV High accuracy for metal or plastic parts High speed 4. System integration and test1. Develop 3D data acquisition systems, including (1) Design coded phase shift projection patterns for purpose of high-performance 3D data acquisition systems in terms of speed, accuracy and measurement range (2) Develop optical engine of 3D data acquisition system excluding camera (contrast: > 400; Image quality: MTF > 0.2 @ 37lps/mm; F number: 3.5) (3) Develop 3D point cloud generation algorithms based on coded phase shift projection patterns (4). 3D data acquisition systems integration Based on the above three technologies, 2 types of 3D data acquisition systems will be delivered for different applications. (a) Narrow-field 3D data acquisition system Field of view (FOV): 100 ~ 340 mm Accuracy: < 0.2mm @1m Speed: 0.5s Measurement range: <1 m (b) Wide-field 3D data acquisition system Field of view (FOV): 248 ~ 828 mm Accuracy: < 0.5mm @1m Speed: 0.5s Measurement range: <1 m 2. Develop 3D object recognition technologies, including (1) Develop fast feature extraction method from 3D data generated by 3D data acquisition systems (2) Develop machine learning based method to identify parts, measure part location and estimate part posture based on the extracted features Based on the above two technologies, 3D object recognition algorithms will be delivered with the following specifications. Recognition time: < 1s Location repeatability: +/- 0.1mm 3. CS deliverables for Customer (2) 1 set prototype of 3D data acquisition system with features: Wide FOV High accuracy for metal or plastic parts High speed 4. System integration and test


    Industrial robots have been adopted in factories worldwide for more than fifty years. In recent years, due to ever increasing wages, diminishing worker supply and shortened product life in consumer electronics, robots are required to be more intelligent for purpose of cooperation with human and even replacement of some workforce. Adoption of intelligent robot enables the enhancement of the flexibility of manufacturing line, in which the lines are able to be efficiently reconfigured for multiple products. To handle such complex tasks in flexible manufacturing, an intelligent robot must be equipped with eyes and brain, which are key capabilities of humanoid cognition. This project aims to develop 3D robot cognition technologies for enabling random bin picking and flexible assembly. To achieve this goal, the project team will develop the following IP protected platform technologies, including (1) robot visual sensing technology by designing 3D data acquisition device and (2) robot recognition technology by designing 3D object recognition algorithm. The global size of industrial robot market is USD 28.93 billion in 2013 and is estimated to reach USD 44.48 billion in 2020. Industrial robots are also elevated as one of strategic sectors in China, according to Made in China 2025 and 13th Five-Year Plan. Hong Kong government has been taking many efforts to promote and support the Innovative and Technology industry. The technologies developed in this project will strengthen the competitiveness of Hong Kong and China companies in this rapid growing industrial robot market.