The

Future

Of Industrial Assembly

POSESTREAM

Cobotic PoseStream is a state-of-the-art deep-learning convolutional neural network designed for real-time object pose estimation. With high precision and accuracy, PoseStream efficiently identifies and locates objects in complex environments, including challenging lighting, reflective or shiny parts, cluttered spaces, and occluded scenes.

Leveraging advanced algorithms, PoseStream determines the optimal grip and approach path for robotic systems, making it ideal for bin picking, precise assembly, and part placement tasks. Its versatile Python API seamlessly integrates with robot controllers, ensuring smooth operations for applications like:

Bin and tray picking
Precision assembly or part placement onto PCBs or trays
Inspection of part placement accuracy

Performance highlights:

Speed: 16 FPS
Accuracy: 1mm in cluttered scenes / 0.3mm in singulated scenes

Whether enhancing automation workflows or improving manufacturing precision, Cobotic PoseStream is engineered to boost productivity and reliability in industrial settings.

FEATURES

Precision Assembly

In-hand Localization After Picking

Collision Free Grip Planning In Real Time

Programming of desired grips

High Speed

APPLICATIONS

Bin Picking

The PoseStream Vision System provides fast, accurate, and dependable performance for bin picking and a variety of other object-handling tasks.

> Learn More

In-hand Positioning

The PoseStream Vision System enables robots to perform in-hand positioning, allowing precise adjustments to objects after they’ve been grasped. The in-hand positioning can be done while the robot is moving to reduce cycle time. This capability ensures optimal orientation for tasks like assembly and placement, enhancing the overall accuracy of the operation. With in-hand positioning, robots can make real-time corrections, improving reliability and reducing the need for regrips or manual intervention.

> Learn More

Inspection

The PoseStream Vision System enables precise inspection during robotic operations, ensuring parts are correctly positioned and oriented. It provides real-time feedback to validate assembly accuracy, helping maintain high-quality standards. With PoseStream's inspection capabilities, the system minimizes errors and reduces the need for manual oversight.

> Learn More

CUSTOMERS

At Audi, PoseStream plays a vital role in a pre-assembly kitting operation, where two robots work in collaboration. The first robot selects wheel bolts from a bin and passes them to the second robot. The second robot then precisely positions and inserts the wheel bolts into an assembly fixture with the correct orientation. PoseStream is integral to this process, ensuring accurate part detection in the bin, planning optimal grasps, locating the part after the handover, and performing the final inspection of the assembly result.

At Robotdalen, PoseStream is utilized in a water testing lab to automate the process of picking four different caps for tubes. The robot handles the testing procedure, with PoseStream accurately locating the caps in bins, planning optimal grasps, and providing in-hand validation for each pick, ensuring precision and efficiency throughout the operation.