Remote Camera Guidance for Delivery Robots at Property Doors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated delivery robots face challenges in navigating to precise delivery points within properties due to insufficient guidance from GPS navigation, especially when encountering obstacles and unknown terrain, as GPS may not accurately direct them to doors or secure areas.
Innovation Solution
Implementing a remote camera-assisted guidance system that uses external cameras to identify and verify robots, determine their capabilities, and select appropriate paths from stored candidate paths based on observed robot movements, providing real-time guidance to ensure safe and accurate delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS navigation is used for robot delivery, then the robot can navigate to general locations, but it cannot accurately direct to specific doors or secure areas within properties
Solution Approach 1:
The patent introduces a remote camera as an intermediary device that captures images of the robot's approach to the property. The camera system acts as a mediator between the robot's GPS navigation and the final delivery point, providing visual information about terrain, obstacles, and door locations that GPS cannot provide. The captured images are transmitted to a server that processes them to generate guidance instructions for the robot.
Solution Approach 2:
The system performs preliminary actions by capturing images of the robot's approach path before the robot reaches the delivery point. The server processes these images in advance to identify terrain features, obstacles, and optimal paths, then provides guidance instructions to the robot. This preliminary analysis allows the robot to navigate complex areas without real-time sensor processing delays.
2Reliability
If the robot navigates autonomously without remote camera assistance, then the system complexity is lower, but the robot cannot identify obstacles and difficult terrain
Solution Approach 1:
The remote camera system serves multiple functions: it captures images of the robot's approach, identifies terrain features and obstacles, determines robot capabilities by analyzing the robot's appearance and movement, and generates guidance instructions. This multi-functional approach increases delivery reliability without requiring separate specialized systems for each function.
Solution Approach 2:
The system uses existing remote cameras (such as doorbell cameras or security cameras) that are already deployed at properties for other purposes. These cameras serve the additional function of assisting robot navigation, eliminating the need to deploy dedicated robot guidance cameras and reducing overall system complexity.
3Ease of operation
If the robot uses GPS navigation only, then the navigation system is simpler, but it is insufficient for guiding to delivery drop-off points near doors
Solution Approach 1:
The navigation process is segmented into two distinct phases: (1) GPS-based navigation to the general property location, and (2) camera-assisted guidance for the final approach to the door. This segmentation allows the system to use the simpler GPS system for the majority of the journey and only activate the more complex camera analysis system when precise door-level navigation is needed.
Solution Approach 2:
The system transitions from two-dimensional GPS coordinate navigation to three-dimensional visual navigation by incorporating camera images that show the physical environment, obstacles, and door locations. This dimensional change from abstract coordinates to visual spatial understanding enables precise guidance to delivery points near doors.
Data Source
AI summary
Methods, systems, and apparatus for remote camera-assisted robot guidance are disclosed. A method includes obtaining images of objects approaching a door of a property; identifying candidate paths to the door based on the images of the objects approaching the door of the property; determining movement capabilities of the objects; storing the candidate paths to the door labeled by the movement capabilities of the objects that took the paths; determining capability information for a robot at the property that indicates movement capabilities of the robot; selecting, from the candidate paths, a path for the robot to take to the door based on the movement capabilities of the robot and the labels of the candidate paths; and providing guidance information to the robot that guides the robot to the door along the selected path.