Cloud Robot Trajectory Caching Over HTTP for Low-Latency Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cloud robotics primarily follows the Platform as a Service (PaaS) approach, which lacks versatility in robot service and control, and existing trajectory or path generation methods do not utilize Web services effectively, leading to a demand for more efficient and scalable robot control solutions.
Innovation Solution
A Web server-based method and system that determines a trajectory for a robot device over a web interface in the cloud by receiving HTTP requests, encoding poses into URI strings, and utilizing a trajectory aware cache to match and generate trajectories efficiently, leveraging existing Web service infrastructure and caching mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If cloud robotics follows the PaaS approach, then processing power and centralized knowledge bases are improved, but service versatility and control flexibility are limited
Solution Approach 1:
The patent implements a universal web service interface that can handle multiple robot control functions through a single standardized HTTP/REST framework. The trajectory generation service accepts generic requests for any robot device and returns standardized trajectory data, making the system adaptable to different robot types and applications without requiring separate control systems for each.
2Measurement precision
If trajectory calculation is performed in real-time without caching, then trajectory accuracy is improved, but computational latency increases
Solution Approach 1:
The system pre-calculates and stores trajectory data in a cache memory based on pose parameters before actual robot execution. When a robot needs to move between poses, the system first checks the cache for pre-computed trajectories matching the current pose parameters, retrieving them instantly without performing full real-time calculations, thus reducing latency while maintaining accuracy.
Solution Approach 2:
The patent replaces the traditional real-time mechanical calculation system with an information-based retrieval system. Instead of computing trajectories through complex real-time mathematical operations, the system substitutes this with database lookups and pattern matching of pose parameters, significantly reducing computational time while preserving trajectory precision.
3Use of energy by moving object
If a trajectory aware cache is implemented, then computational resource usage is reduced, but system complexity increases
Solution Approach 1:
The system creates simplified copies of trajectory data in cache memory that replicate the essential pose-parameter relationships without storing full computational models. These cached trajectory representations are lightweight data structures that can be quickly retrieved and applied, reducing the need for complex real-time computations while maintaining system functionality.
4Measurement precision
If custom robot control protocols are developed, then control precision is improved, but ease of operation and integration difficulty increase
Solution Approach 1:
The patent adopts universal HTTP and REST protocols for robot control, allowing the system to communicate with any robot device using standard web technologies. This universal interface maintains control precision through structured pose parameter representations and trajectory data formats while dramatically improving ease of operation, as developers can integrate robot control using familiar web development skills without learning proprietary protocols.
Data Source
AI summary
The present disclosure relates to a Web server (104, 60, 70) and a method therein of determining a trajectory for controlling of a robot device over a cloud interface. From a URI-encoded HTTP request (402) for the trajectory between a first pose and a second pose of the robot device, it is determined (S112, 408) the length of matching between cached trajectories and the trajectory requested. The longest length of matching is compared (S114, 410) to a minimal matching length (406), and if the longest length is longer than the minimal matching length, a HTTP response is sent (S122, 412) comprising the trajectory being determined. If the longest length is shorter than the minimal matching length, a HTTP request to calculate is sent (S116, 414). Currently available web service architecture can be reused, and easily up-scaled.


