3D Machine Vision System | CoPick3D
Automated Solution
3D Machine Vision for any kind of manufacturing
The CoPick3D vision platform provides a clear path from pilot cells to 24/7 robotic picking by integrating an industrial camera with fast, wide-area 3D capture to reduce system complexity and cost. It delivers accurate guidance accuracy and collects millions of 3D points within seconds. Deploy quickly in constrained footprints and fine-tune precision for small or delicate parts, then scale to high-throughput mixed bins and extend reach for deep bins or bulky components. Improve your production process with single solution, CoPick3D.
What 3D Machine Vision enables - 4 examples of use case
Automated Tire–Wheel Mounting | Deep-Learning Visual Inspection
FineLocalizer—optimized for CoPick3D—recognizes holes and studs and applies geometric constraints (e.g., square, pentagon, hexagon) to estimate full hub pose from as few as three features. With CoPick3D’s color-texture imaging, the same station executes automatic mounting and heterogeneous visual inspection in a single pass, consolidating multiple vision systems and reducing equipment cost and footprint. Deep-learning recognition then classifies tires and wheels across variants and enforces vehicle-specific sequences. The result is error-free, high-precision installation.
Automated Handling of Non-Uniform Wiring Connectors
CoPick3D, paired with FineLocalizer robot-guidance software, automates one of assembly’s hardest jobs: attaching and detaching non-uniform wire harnesses and connectors with repeatable accuracy. In a representative EV use case, a test-module cable with three coupled connectors is mated in the prescribed order and within tight cycle-time limits. Replacing manual steps with robotic gripping and guided insertion, the system maintains high positional precision while delivering a step-change in throughput.
Bolt Picking System
With CoPick3D, a single structured-light scan identifies up to a user-defined number of bolts in a randomly stacked bin. The robot is then guided to pick them—sequentially and precisely—for the nutrunner magazine. User Preference-Based Bin Picking sustains performance under uncertainty—thread damage, part shift, occlusions—so output remains consistent. Compared with drum-type feeders, the workcell is markedly more flexible: it handles diverse bolt types, rejects defects, and eliminates dummy-capacity constraints.
Automated Windshield & Sunroof Installation
Built on CoPick3D, this 3D machine-vision solution pairs with FineLocalizer to correct vehicle-body misalignment with high precision—even when bodies arrive slightly offset. It detects real-time errors from both the glass-alignment unit and the gripper, and guides the robot by referencing the relative pose of the body and glass simultaneously. A single system performs placement and in-line quality inspection, stabilizing process quality while enabling continuous quality monitoring.
Tokyo Boeki North America supports your efficient production process
See the catalog for specifications
We will manage these personal information under appropriate safety measures and we will not disclose / offer to third parties without your consent in advance. Please refer to our company privacy policy.
Explore related products
