AR Enhanced Robotic Animation Feedback Loop

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current animation tools fail to provide a lifelike view of interactive robotic characters due to the incomplete feedback loop between human and robot interactions, as standard simulators do not fully account for the dynamic movement of sensors like cameras on robots, affecting the animation and overall system performance.

Innovation Solution

The implementation of an augmented reality (AR) enhanced interactive robotic animation system that integrates perception and animation software, using machine learning models to process environmental data from sensors and generate performative actions, including AR effects, to create a seamless human-robot interaction simulation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If standard animation tools are used to animate robotic characters, then the animation can be created, but the feedback loop between human and robot actions is incomplete, resulting in inaccurate representation of lifelike interactions

Engineering Contradiction:
Improveaccuracy of interaction simulationVSAvoidcomplexity of simulation system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent creates a virtual copy of the physical robot with identical sensor placements and movement characteristics. This virtual robot copies the sensor data acquisition and processing from the physical robot, allowing the animation system to accurately simulate how the robot perceives and reacts to human actions, thereby completing the feedback loop without requiring the full physical hardware

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent introduces an intermediary simulation environment that mediates between the physical robot and the animation system. This intermediary layer processes sensor data from the physical robot and generates corresponding visual feedback, creating a closed-loop system where human actions are accurately reflected in the animated representation and vice versa

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If a simulator with recorded human inputs is used, then some interaction feedback is provided, but the sensor movement dynamics are not captured, leading to incomplete feedback loop

Engineering Contradiction:
Improveinteraction feedback capabilityVSAvoidsensor perspective accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent makes the virtual sensors dynamic by coupling their movement to the virtual robot's body movements. As the virtual robot moves its head, eyes, or other sensor-bearing components, the virtual sensors move accordingly, accurately capturing the changing perspectives and measurement conditions that occur during natural robot-human interaction

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS11747890B2Augmented reality enhanced interactive robotic animation
Publication Date: 2023.09.05 DISNEY ENTERPRISES INC
  • US11747890B2 patent drawing
  • US11747890B2 patent drawing
  • US11747890B2 patent drawing

AI summary

A system includes a computing platform having processing hardware, one or more sensor(s), and a memory storing perception software and animation software, as well as an augmented reality (AR) headset including another processing hardware. The computing platform is configured to obtain an environmental data using the sensor(s), determine the location and orientation of the AR headset using the perception software and the environmental data, identify an action for execution by the computing platform, using the animation software and the location and orientation of the AR headset, and transmit performative data corresponding to the identified action to the AR headset. The AR headset is configured to receive, from the computing platform, the performative data, and to render at least one AR effect corresponding to the identified action using the performative data.