Elevator Car Capacity Control Using Image-Based Passenger Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Elevator systems lack the ability to accurately determine the number of additional passengers that can board a car, leading to inefficient car assignments and potential overcrowding, especially when groups of passengers are involved, resulting in wasteful stops and reduced system efficiency.
Innovation Solution
The method involves using image processing units to capture real-time images of the elevator car and hallway, creating graphic models of passengers to determine available space, and counting standard-sized passenger models to assess how many more passengers can fit, with additional features for detecting groups of passengers, ensuring accurate car assignment based on actual capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If car assignments are made using destination entry systems without real-time capacity information, then car assignment algorithms can operate, but passengers may be assigned to cars that are already full or nearly full, causing wasteful stops and reduced system efficiency
Solution Approach 1:
The system performs preliminary actions by continuously capturing images of the elevator car interior and calculating available capacity before passengers board. The image processing unit creates passenger models and determines how many more passengers can fit, providing this information to the car assignment algorithm in advance to make informed decisions.
Solution Approach 2:
The system implements feedback by continuously monitoring the actual number of passengers in the elevator car using image processing and comparing it with the car's maximum capacity. This real-time capacity information is fed back to the car assignment algorithm, allowing dynamic adjustment of assignments based on actual rather than theoretical availability.
2Measurement precision
If the elevator system uses conventional load-based capacity assessment, then capacity calculations can be performed, but the actual physical space available for passengers cannot be accurately determined, leading to incorrect car assignments
Solution Approach 1:
The system replaces traditional mechanical or sensor-based load measurement with an optical imaging system. Instead of using weight sensors or complex mechanical devices to measure capacity, the system uses image processing units with cameras to visually assess the elevator car interior and calculate available space using passenger models.
Solution Approach 2:
The system creates graphic passenger models as simplified representations or copies of actual passengers. These standard-sized passenger models are placed in the image area to simulate occupancy and determine how many more passengers can fit, providing an accurate visual measurement of available capacity without requiring physical measurement devices.
3Productivity
If destination entry systems assign cars without considering group dynamics, then individual passengers can be served, but groups of passengers wishing to travel together cannot be properly accommodated, causing additional stops and reduced efficiency
Solution Approach 1:
The image processing system serves multiple functions: it counts individual passengers, detects groups of passengers, and calculates available capacity. By making the passenger detection system multi-functional, the system can adaptively serve both individual passengers and groups without requiring separate detection mechanisms, improving overall system versatility.
Data Source
Figure 1
Figure 2
Figure 3a~3d
AI summary
A method for controlling an elevator system comprises obtaining images of the inside of an elevator car (101), detecting passengers in the images (102), creating graphic passenger models of the detected passengers (103) from the images and determining the number of passengers who may additionally board the elevator car using the passenger models (109). The passenger models may reflect the actual size of the passengers.