Robot Graphical User Interface for Semi-Autonomous Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot training methods, particularly for autonomous control, are costly and impractical due to the need for physical rigs and 1:1 pilot-to-robot ratios, and teleoperation systems require complex interfaces that limit accessibility.
Innovation Solution
A control system utilizing a cognitive architecture with a graphical user interface that allows a human operator to control a robot semi-autonomously by selecting instructions from a pre-determined set based on sensor data, enabling training in both physical and simulated environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If teleoperation systems are used to control robots, then the robot can be operated by a human pilot, but the interface becomes elaborate and complicated requiring full attention from the pilot
Solution Approach 1:
The patent uses a simulated robot in a virtual environment as a copy of the physical robot. The simulated robot receives the same sensor data and processes it through the same cognitive architecture, allowing pilots to train on a simplified virtual representation without needing complex physical pilot rigs or immersive interfaces.
Solution Approach 2:
The patent introduces a simulated robot as an intermediary between the pilot and the physical robot. The simulated robot processes sensor data and generates control commands that can be transferred to the physical robot, allowing the pilot to operate through a simplified graphical interface rather than direct physical control.
2Reliability
If physical robots are used for training autonomous control, then the robot can learn through repeated physical tasks, but the costs and materials become prohibitive
Solution Approach 1:
The patent creates a simulated robot that copies the physical robot's sensor suite and cognitive architecture. This virtual copy can be trained repeatedly through simulation without the need for physical materials, reducing training costs while maintaining training effectiveness through equivalent sensor data and control workflows.
Solution Approach 2:
The patent performs training actions preliminarily in the simulated environment before transferring to the physical robot. The simulated robot can undergo extensive training and iteration in advance, allowing the physical robot to benefit from pre-acquired knowledge and reducing the need for expensive physical training repetitions.
3Ease of operation
If a pilot rig is used for training robots, then the robot can be controlled analogously, but the hardware requirements and space needs become significant
Solution Approach 1:
The patent replaces the physical pilot rig with a virtual simulation environment. The simulated robot receives sensor data from physical sensors and processes it through the same cognitive architecture, providing analogous control experience without requiring physical pilot rig hardware or specialized equipment.
Solution Approach 2:
The patent substitutes the mechanical pilot rig system with a software-based simulation system. Instead of physical haptic feedback and mechanical interfaces, the system uses virtual representations and graphical user interfaces to provide control, eliminating the need for complex mechanical hardware.
4Loss of time
If intermediate level training is used, then costs and time are reduced, but the training must be as similar to fully autonomous control as possible to be effective
Solution Approach 1:
The patent uses a simulated robot that copies the physical robot's sensor data processing and cognitive architecture. This ensures that intermediate training through the simulation is as similar to fully autonomous control as possible, maintaining training effectiveness while reducing time and cost through virtual repetition.
Data Source
AI summary
Control systems for controlling operation of a robot, as well as computer program products and methods thereof are provided herein, The control system comprises a robot, including a plurality of sensors configured to convert information from the environment and the robot into sensor data and a plurality of actuators; a cognitive architecture control system, communicatively coupled to the robot, configured to control the robot, to receive the sensor data, to generate a robot-egocentric model of the environment from the sensor data, and to output autonomous actuator data to the plurality of actuators of the robot, based on at least one instruction; and a graphical user interface configured to display a graphical representation of the robot-egocentric model to a human operator and enable the human operator to select the at least one instruction from a pre-determined instruction set, based on the robot-egocentric model, to control the robot semi-autonomously.