Mobile Robot Hazard Detection With Bandwidth-Aware Event Upload
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Retail environments face significant costs due to pedestrian injuries from walking hazards, with expenditures for safety and shopping experience improvements often disassociated, leading to exorbitant costs for providers.
Innovation Solution
A mobile robot equipped with image and infrared sensors, onboard processing, and a cloud-connected system for real-time hazard detection and path modification, combined with distributed processing architecture for efficient data management and action, to prevent pedestrian injuries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all hazard detection data is stored and processed locally on the mobile robot, then real-time hazard detection accuracy is improved, but device complexity and energy consumption increase
Solution Approach 1:
The system divides processing tasks between the mobile robot (real-time hazard detection and path modification) and the cloud service (data storage and analytics). This segmentation allows the robot to maintain simple onboard processing while achieving accurate hazard detection through cloud-supported analytics.
Solution Approach 2:
The cloud service acts as an intermediary between the mobile robot and the hazard detection data. The robot sends data to the cloud and receives analytics results, allowing complex processing to occur remotely while the robot maintains real-time control capabilities.
2Speed
If real-time hazard detection and path modification is performed using onboard processing only, then response time is improved, but loss of energy increases due to continuous local computation
Solution Approach 1:
Processing tasks are segmented between the mobile robot (time-critical real-time detection and control) and the cloud service (non-time-critical storage and analytics). This allows the robot to consume less energy by performing only essential real-time processing locally.
Solution Approach 2:
The robot performs only the minimum necessary processing locally (real-time hazard detection and path modification) while offloading more comprehensive analytics to the cloud. This partial action approach optimizes energy consumption while maintaining safety-critical response times.
3Measurement precision
If comprehensive hazard detection data is continuously uploaded to cloud storage, then analytics accuracy is improved, but bandwidth consumption increases
Solution Approach 1:
The system extracts and uploads only the essential event records (timestamp, robot pose, hazard class, confidence value) to the cloud, rather than continuously uploading all raw sensor data. This extraction approach maintains analytics accuracy while significantly reducing bandwidth consumption.
Solution Approach 2:
Instead of continuously uploading all detection data, the system selectively uploads only relevant event records that meet certain criteria. This partial upload strategy provides sufficient data for accurate analytics while minimizing bandwidth usage and energy consumption.
Data Source
AI summary
Systems and methods include: a mobile robot comprising a drive system, at least one image sensor; at least one infrared sensor, and an onboard processing system having at least one processor and a memory storing instructions; a store gateway communicatively coupled to the mobile robot and configured to receive an event record under a bandwidth-aware policy, to provide at least aspects of the event record for upload for storage to a cloud service, and to receive updates from the cloud service for the mobile robot; and an analytics module at the cloud service communicatively coupled to the store gateway and configured to perform analytics of uploaded ones of the event records.