Robot interaction with human co-workers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robots in fast-paced environments like restaurant kitchens struggle to manipulate unpredictable and shape-changing food ingredients without altering their texture and taste, and existing collision avoidance methods are inadequate for safe human-robot collaboration.
Innovation Solution
A robot system that uses computer vision and neural networks to predict the motion of obstacles, allowing it to generate motion plans that avoid collisions with humans and objects by extending the time horizon of sensed objects in the environment, ensuring safe and efficient collaboration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional material handling methods (auger, conveyor, suction mechanisms) are used to move food ingredients, then the robot can transport materials, but the friction, stiction, and viscosity of foods cause these mechanisms to become clogged and soiled, and they impart forces on the foodstuffs which alter their texture, consistency, and taste-profile in unappetizing ways
Solution Approach 1:
The patent replaces traditional mechanical material handling systems (augers, conveyors, suction mechanisms) with a robotic arm that uses controlled grasping and positioning to manipulate food ingredients. This substitution eliminates the harmful friction and viscous forces that altered food texture and taste, while maintaining the ability to transport and position ingredients accurately in the food preparation environment
2Productivity
If the robot moves quickly to maintain productivity in fast-paced environments, then task completion speed increases, but the robot cannot adequately detect and respond to moving obstacles such as human chefs, leading to potential collisions
Solution Approach 1:
The patent implements preliminary action by continuously predicting the future positions of moving obstacles (particularly human chefs) using learned motion patterns before collisions can occur. This allows the robot to proactively adjust its trajectory and speed to avoid collisions, maintaining both high productivity and safety in fast-paced kitchen environments
Solution Approach 2:
The system employs feedback mechanisms by continuously monitoring the positions and motions of obstacles, comparing predicted positions with actual positions, and adjusting the robot's motion plan in real-time. This closed-loop control enables the robot to respond dynamically to changing conditions while maintaining high-speed operation
3Reliability
If the robot extends the time horizon of sensed objects into the future to predict motion and improve safety, then collision avoidance improves, but the complexity of motion planning and computation increases
Solution Approach 1:
The patent applies parameter changes by representing obstacle positions and motions in a transformed coordinate system that simplifies prediction and collision detection. By changing the parameters of how spatial and temporal relationships are represented, the system achieves accurate long-term motion prediction without proportionally increasing computational complexity
Solution Approach 2:
The system creates simplified copies or models of the kitchen environment and obstacle motions that capture essential dynamics without requiring full physical simulation. These computational models allow the robot to predict future states efficiently while maintaining safety, reducing the complexity burden of extended time horizon planning
Data Source
AI summary
Embodiments provide functionality to prevent collisions between robots and objects. An example embodiment detects a type and a location of an object based on a camera image of the object, where the image has a reference frame. Motion of the object is then predicted based on at least one of: the detected type of the object, the detected location of the object, and a model of object motion. To continue, a motion plan for the robot is generated that avoids having the robot collide with the object based on the predicted motion of the object and a transformation between the reference frame of the image and a reference frame of the robot. The robot can be controlled to move in accordance with the motion plan or a signal can be generated that controls the robot to operate in accordance with the motion plan.