18126900. COLLISION DETECTION FOR OBJECT REARRANGEMENT USING A 3D SCENE REPRESENTATION simplified abstract (NVIDIA Corporation)

From WikiPatents
Jump to navigation Jump to search

COLLISION DETECTION FOR OBJECT REARRANGEMENT USING A 3D SCENE REPRESENTATION

Organization Name

NVIDIA Corporation

Inventor(s)

Adithyavairavan Murali of Seattle WA (US)

Arsalan Mousavian of Seattle WA (US)

Clemens Eppner of Seattle WA (US)

Adam Fishman of Seattle WA (US)

Dieter Fox of Seattle WA (US)

COLLISION DETECTION FOR OBJECT REARRANGEMENT USING A 3D SCENE REPRESENTATION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18126900 titled 'COLLISION DETECTION FOR OBJECT REARRANGEMENT USING A 3D SCENE REPRESENTATION

Simplified Explanation

The abstract of the patent application describes a method for collision detection in object rearrangement tasks using a three-dimensional scene representation.

  • The innovation involves using a three-dimensional scene representation for collision detection in object rearrangement tasks.
  • The method includes a robot manipulator picking up a target object and placing it in a target location, such as a shelf or cabinet.
  • The technology requires skills in picking, placing, and generating complex collision-free motions in cluttered environments.

Potential Applications

The technology can be applied in industries such as warehouse automation, manufacturing, and logistics for efficient object rearrangement tasks.

Problems Solved

1. Ensures collision-free object rearrangement in cluttered environments. 2. Improves the efficiency and accuracy of robotic tasks involving object manipulation.

Benefits

1. Increases productivity by automating object rearrangement tasks. 2. Reduces the risk of collisions and damage to objects or the environment. 3. Enhances the precision and reliability of robotic manipulations.

Potential Commercial Applications

Optimizing Object Rearrangement Tasks in Manufacturing and Logistics

Possible Prior Art

Prior art in collision detection and object manipulation in robotics may include research papers, patents, or existing technologies that address similar challenges in object rearrangement tasks.

Unanswered Questions

How does the three-dimensional scene representation improve collision detection in object rearrangement tasks?

The three-dimensional scene representation allows for a more accurate and detailed understanding of the environment, enabling the robot to detect potential collisions with other objects or obstacles.

What are the limitations of the current technology in object rearrangement tasks?

The current technology may face challenges in handling highly dynamic environments, complex object shapes, or unpredictable obstacles that could affect the accuracy and efficiency of object rearrangement tasks.


Original Abstract Submitted

One common robotic task is the rearrangement of physical objects situated in an environment. This typically involves a robot manipulator picking up a target object and placing the target object in some target location, such as a shelf, cabinet or cubby, and requires the skills of picking, placing and generating complex collision-free motions, oftentimes in a cluttered environment. The present disclosure provides collision detection for object rearrangement using a three-dimensional (3D) scene representation.