Tiered Robot Compute Architecture for Low-Latency Autonomous Operations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial robotics face challenges in performing autonomous operations due to high computational complexity and power consumption, which can be mitigated by distributing compute resources between robots and servers, but require minimal network latency and maximal security.
Innovation Solution
A tiered network architecture is implemented, where a top-tier cloud server acts as the truth source and an on-premises fog or edge server executes computationally intensive processes for robots, with robots performing latency-sensitive operations locally and communicating with servers for data processing and execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If compute resources are implemented on the robot to perform autonomous operations, then operational capability and automation are improved, but device complexity and power consumption increase
Solution Approach 1:
The system segments compute resources into two distinct layers: edge compute resources embedded in the robot for autonomous operations, and cloud compute resources for complex processing. This segmentation allows the robot to perform basic autonomous functions independently while offloading heavier computational tasks to the cloud, thereby reducing on-robot hardware complexity while maintaining automation capability.
Solution Approach 2:
The patent introduces an edge server as an intermediary between the robot and cloud infrastructure. This edge server acts as a mediator that provides additional compute resources when needed, allowing the robot to maintain simpler onboard hardware while still accessing powerful computational capabilities through the edge-server-cloud hierarchy.
2Extent of automation
If compute resources are implemented on the robot, then autonomous operation capability is improved, but power consumption increases
Solution Approach 1:
The compute workload is segmented between edge and cloud resources. The robot's edge compute handles only essential autonomous operations with low power consumption, while computationally intensive tasks are offloaded to cloud data centers with unlimited power supply, thereby maintaining automation capability while minimizing robot power consumption.
Solution Approach 2:
The system implements partial local execution and partial cloud execution. Only the minimum necessary compute operations are performed on the robot to maintain autonomy, while excessive or non-critical computational tasks are executed in the cloud, optimizing the balance between autonomous capability and power consumption.
3Adaptability or versatility
If computationally intensive processes are executed on the robot, then operational flexibility is improved, but latency increases for time-sensitive operations
Solution Approach 1:
The system segments operations into two categories: latency-sensitive operations executed locally on the robot's edge compute, and non-time-critical operations executed in the cloud. This segmentation ensures that time-sensitive autonomous operations maintain low latency while still utilizing cloud resources for flexible, computationally intensive tasks.
Solution Approach 2:
Different parts of the system have different computational qualities assigned to them. The robot's edge compute is optimized for low-latency local operations, while cloud data centers provide high-capacity compute for flexible but less time-sensitive tasks. This local quality differentiation resolves the contradiction between operational flexibility and latency.
Data Source
AI summary
The present disclosure relates to a multi-tiered computing environment for balancing compute resources in support of robot operations. In an example, a robot is tasked with performing an operation associated with an airplane having an airplane model. To do so, the robot may need another operation that is computationally complex to be performed. An on-premises server can execute a process that corresponds to this computationally-complex operation based on sensor data of the robot and can output the resulting data to the robot. Next, the robot can use the resulting data to execute another process corresponding to its operation and can indicate performance of this operation to the on-premises network. The on-premises network can send the indication about the operation performance to a top-tier server that is also associated with the airplane model.