Industrial IoT Orchestration With Edge AI for Multi-Robot Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current industrial robot systems face challenges with high hardware and maintenance costs, difficulty in sharing sensors, complex collaboration logic across multiple robots, and scalability issues due to the centralized control architecture and the need for powerful hardware to support advanced AI models.
Innovation Solution
The system leverages a cloud foundation, including a 5G-empowered cloud, to enable IoT sensors, AI models, and robots to collaborate effectively by extracting robot software functions into containers running on an edge cloud, separating mechanical control functions, and deploying AI models and sensors as containers on the cloud, with an orchestration service managing collaboration across multiple devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If centralized control architecture is used with powerful hardware to support advanced AI models, then manufacturing precision and productivity are improved, but device complexity and hardware costs increase
Solution Approach 1:
The patent segments the centralized control architecture into distributed edge computing nodes. Each robot or device group has its own edge computing device that runs local AI models, dividing the previously centralized computational load into multiple independent segments. This reduces the complexity of any single device while maintaining overall system intelligence through coordinated operation of multiple edge nodes.
Solution Approach 2:
The patent extracts AI model execution from the centralized cloud infrastructure and places it at the edge computing devices located at or near the robots. This extraction removes the dependency on powerful centralized hardware for real-time control, allowing simpler local devices to perform sophisticated AI tasks independently while the cloud handles only non-real-time functions like model training and updates.
2Productivity
If centralized control architecture with powerful hardware is used, then productivity is improved, but hardware costs and maintenance costs increase
Solution Approach 1:
The patent implements local quality by deploying AI capabilities directly at the edge computing devices near each robot rather than requiring all robots to connect to powerful centralized hardware. Each edge device is configured with appropriate computational resources for its specific tasks, optimizing cost-efficiency by matching hardware capabilities to actual local needs rather than providing every node with high-end equipment.
Solution Approach 2:
The patent introduces edge computing devices as intermediaries between the robots and the centralized cloud infrastructure. These edge devices handle real-time AI inference and local control tasks, reducing the need for expensive high-performance hardware at every robot station. The edge layer mediates between simple robot controllers and the powerful cloud, allowing cost-effective hardware configurations while maintaining high productivity through efficient local processing.
3Measurement precision
If sensors are physically attached to robots, then measurement precision is improved, but ease of operation and scalability worsen due to difficulty in sharing sensors
Solution Approach 1:
The patent creates virtual copies of sensor data through the digital twin framework. Instead of requiring physical sensor attachment to each robot for data access, the system generates virtual representations of sensor information that can be replicated and shared across multiple robots and applications. This copying approach maintains measurement precision by preserving accurate sensor data while enabling unlimited sharing and reuse without additional physical hardware.
Solution Approach 2:
The patent implements universality by creating a shared sensor data infrastructure where a single physical sensor can serve multiple robots and functions simultaneously. Through the edge computing platform and digital twin technology, sensor data is made universally accessible to any component that needs it, transforming previously dedicated one-to-one sensor-robot connections into many-to-many relationships that enhance adaptability and sensor utilization efficiency.
Data Source
AI summary
In an approach to AI empowered factory automation using an industrial IoT, responsive to receiving a new production task, production requirements are input into an AI engine. Equipment and software containers are selected based on the AI model. An orchestration service is created, where the orchestration service collaborates the equipment and the software containers.