A real-time robot guidance system from Inbolt is reported to enable flexible and low-cost bin picking for unstructured environments.
DETROIT—A next-generation bin picking system recently introduced by Inbolt is designed to bring human-like adaptability to industrial robotics.
Built for fully unstructured environments, the system enables robots to identify, grasp, and place parts with high flexibility and accuracy, even when objects are randomly positioned or partially hidden, the company said in a release.
Inbolt Intelligent Bin Picking brings AI 3D vision directly onto the robot arm. In doing so, it enables robots to pick parts from any bin with up to a 95 percent success rate and sub-one-second cycle times, the company stated. (Image: Inbolt/Business Wire)
According to Inbolt, traditional bin picking systems rely on fixed overhead, long-range 3D cameras, complex calibration, and pre-calculated grasp points. Such systems can be expensive, rigid, and easily disrupted when bins move, parts aren’t detected, or pre-defined pick points aren’t accessible.
Inbolt, a developer of real-time vision-guidance systems for robots, developed an approach that
turns this model on its head. The robot uses a 3D camera mounted directly on the robot arm, powered by Inbolt’s proprietary AI, to continuously perceive, understand, and adapt in real time.
“The AI delivers an infinite number of grasp strategies, removing the need for a perfect grasp,” the release stated. “This provides a faster, cheaper, and more flexible solution that achieves less than 1 second per pick and up to 95 percent success rates in live manufacturing production.”
The bin picking system uses a process inspired by human behavior. According to the release, the robot identifies any pickable side, just like a human. Once the object is gripped, the robot analyzes and locates it. During motion, the AI is said to continuously refine the robot’s trajectory for accurate placement, a capability known as in-hand localization.
“This closed-loop process allows the robot to adapt instantly to part variability and bin movement, achieving a level of robustness never before seen in traditional bin picking systems,” the company stated in the release.
Inbolt’s on-arm camera architecture is said to eliminate the need for multiple fixed cameras and expensive high-resolution setups. Manufacturers can deploy the same robot across different bins and configurations, significantly lowering hardware costs and setup time.
“Traditional bin picking systems are too rigid for real factory conditions,” said Albane Dersy, COO of Inbolt, in the release. “We designed our solution to adapt in real time, able to see, grasp, and adjust the way a human would. That level of flexibility is what manufacturers need to reach truly autonomous production.”
Inbolt’s unstructured bin picking system is currently running across more than five different factories, the company said in the release. Across these sites, the technology is said to have “consistently delivered high uptime and throughput, proving its robustness in real-world industrial conditions.”
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.
Its unique architecture is said to minimize computational load while maximizing robustness, ensuring consistent performance across different use cases and part geometries.