Robotic Vehicle Pose Validation for Infrastructure Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous mobile robots (AMRs) face challenges in determining a pose to interact with payloads on fixed infrastructure without colliding, as current methods rely on manually trained poses that do not adjust for the actual orientation of the payload, leading to potential collisions with obstacles.
Innovation Solution
The AMR is equipped with sensors to acquire real-time data, processing it to generate exclusion regions or volumes to determine if a desired pose will result in collisions, allowing it to adjust its pose dynamically to avoid infrastructure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manually trained poses are used for AMR to interact with payloads on fixed infrastructure, then the AMR can achieve a predetermined position, but it cannot adjust for actual payload orientation leading to potential collisions
Solution Approach 1:
The patent transforms the static, manually trained pose into a dynamic, validated pose through real-time sensor data processing. The pose validation system dynamically adjusts the AMR's target pose based on actual infrastructure geometry detected by sensors, making the pose adaptive rather than fixed. This resolves the contradiction by maintaining precision through validation while achieving adaptability through real-time adjustment.
Solution Approach 2:
The patent implements a feedback loop where sensor data about the actual infrastructure and payload orientation is processed to validate and potentially adjust the manually trained pose. The system compares expected pose parameters with real-time sensor measurements and modifies the target pose accordingly. This feedback mechanism enables the AMR to adapt to actual conditions while maintaining the precision of manually trained positions.
2Ease of operation
If the AMR uses protruding chassis parts like outriggers to reach payloads, then it can access harder-to-reach positions, but it may collide with fixed infrastructure
Solution Approach 1:
The patent applies preliminary action by validating the pose before the AMR executes the movement to reach the payload. The pose validation system processes sensor data and determines collision risk in advance, allowing the AMR to adjust its approach or select alternative poses before attempting to reach difficult-to-access positions with protruding parts like outriggers. This prevents collisions while maintaining accessibility.
Solution Approach 2:
The patent implements preliminary anti-action by identifying potential collision hazards before the AMR attempts to reach payloads with protruding chassis parts. The pose validation system detects infrastructure that would be collided with and prevents the harmful action by adjusting the pose or blocking the approach. This allows the AMR to maintain ease of operation for accessing difficult positions while preventing collisions through advance hazard identification.
3Reliability
If the AMR validates pose using sensor data processing, then it can avoid collisions with infrastructure, but it requires additional computational processing time
Solution Approach 1:
The patent extracts only the critical information needed for collision detection from the full sensor data set. The pose validation system processes sensor data to identify specific geometric features and obstruction regions relevant to the intended pose, rather than analyzing all sensor data comprehensively. This extraction approach maintains high reliability for collision avoidance while reducing computational time by focusing only on essential validation parameters.
Data Source
AI summary
A robotic vehicle comprising a chassis and a manipulatable payload engagement portion, at least one sensor configured to acquire real-time sensor data, a pose validation system comprising computer program code executable by at least one processor to evaluate the sensor data to: determine if a goal pose of the robotic vehicle will result in a collision with infrastructure upon which the object is located when the engagement portion engages the object. If a potential collision is detected, the pose validation system can generate a signal to adjust the robotic vehicle's pose to avoid the collision. A corresponding method is also provided.