Robot interaction with human co-workers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic systems in food-service settings face challenges in safely collaborating with humans due to unpredictable ingredient arrangements and the need to avoid collisions with moving obstacles, as traditional material handling methods often result in texture and taste alterations of food, and are ill-suited for cooked foods, leading to inefficiencies and safety concerns.
Innovation Solution
A robotic system equipped with an articulated arm that uses computer vision and neural networks to detect and predict the motion of obstacles, including humans, allowing it to generate motion plans that avoid collisions and ensure safe collaboration by extending the time horizon of sensed objects and considering future motion, thereby preventing collisions and allowing efficient task execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional material handling methods (auger, conveyor, suction mechanisms) are used to move foodstuffs, then material can be transported, but the friction, stiction, and viscosity cause mechanisms to become clogged and soiled, and forces imparted alter 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 system that uses controlled mechanical manipulation to transfer foodstuffs. This substitution eliminates the harmful friction and viscosity-related issues caused by traditional mechanisms while maintaining material transport capability through precise robotic control.
Solution Approach 2:
The patent changes the operational parameters by using a robotic arm with controlled force application, varying speeds, and precise positioning to handle foodstuffs. This allows adaptation to different food textures and viscosities without the clogging and soiling problems of traditional fixed-parameter mechanisms, thereby preserving food quality during transport.
2Productivity
If robots operate in high-traffic, fast-paced environments like restaurant kitchens, then task execution speed increases, but collision with moving obstacles (humans, objects) becomes more likely
Solution Approach 1:
The patent implements preliminary action by equipping the robot with sensors and vision systems that detect obstacles and predict their motion trajectories in advance. This allows the robot to plan and adjust its path proactively before collisions occur, enabling high-speed operation in dynamic kitchen environments while maintaining safety through predictive obstacle avoidance.
Solution Approach 2:
The patent employs feedback mechanisms where sensors continuously monitor the environment for moving obstacles, and this real-time information feeds back to the robot's control system. The system processes this feedback and dynamically adjusts the robot's motion plan, allowing fast task execution while reliably avoiding collisions through continuous environmental awareness and adaptive response.
3Reliability
If the robot extends the time horizon of sensed objects into the future to predict motion, then safety of operation increases, but computational complexity and processing time increase
Solution Approach 1:
The patent applies partial action by predicting motion only for detected obstacles within relevant time horizons, rather than computing all possible future states. This selective prediction approach maintains high safety through accurate forecasting of near-term obstacle trajectories while reducing computational complexity by focusing resources on the most critical predictions for immediate collision avoidance.
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.