This news came to me touted as an industry first. Actually, I’ve already written about a similar application. That does not denigrate the accomplishment. It only speaks to marketing. AI has been a part of vision systems as long as I can remember. AI as an automation tool continues to become more powerful and useful. Just go past hype, as I’ve written before, and look at the usefulness (or hype).
My first brush with vision systems was long ago. I told prospects that if we could fixture the part, we could check it. Speed and precision in cameras and processors continually improved. Now we have a further advance—flexible and low-cost bin picking for unstructured environments, achieving 95% success rates and sub-1-second pick cycles in live automotive production.
Inbolt announced the launch of its next-generation bin picking solution designed to bring human-like adaptability to industrial robotics. Built for fully unstructured environments, this breakthrough enables robots to identify, grasp, and place parts with unmatched flexibility and accuracy, even when objects are randomly positioned or partially hidden.
Using a 3D camera mounted directly on the robot arm, powered by Inbolt’s proprietary AI, the robot continuously perceives, understands, and adapts in real time. The AI delivers an infinite number of grasp strategies, removing the need for a perfect grasp. This provides a faster, cheaper, and more flexible solution that achieves less than 1 second per pick and up to 95% success rates in live manufacturing production.
How the solution operates inspired by human behavior.
- Pick like a human — The robot identifies any pickable side, just like a human.
- See in hand — Once the object is gripped, the robot analyzes and locates it.
- Adjust to place — During motion, the AI continuously refines the robot’s trajectory for accurate placement, a capability known as in-hand localization.
Key benefits include:
- Flexibility: Works with bins of any size or position, even if they move during operation.
- Speed: Average processing time per pick under 1 second.
- Reliability: Up to 95% successful pick rates in production environments.
- Affordability: One camera per robot, not per bin, and no need for complex overhead installations.
- Scalability: Fast integration across multiple stations or factories.
- Production-proven across plants
The system runs on NVIDIA’s hardware platform and leverages Inbolt’s proprietary AI robot guidance models, which enable real-time pose estimation and continuous trajectory correction. This unique architecture minimizes computational load while maximizing robustness, ensuring consistent performance across different use cases and part geometries.
