Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Real-time autonomous crowd control system via detection of crowd anomalies from peer-to-peer devices

IP.com Disclosure Number: IPCOM000246067D
Original Publication Date: 2016-May-02
Included in the Prior Art Database: 2016-May-02
Document File: 7 page(s) / 541K

Publishing Venue

Motorola

Related People

Lee, Kuan Heng: INVENTOR [+5]

Abstract

Recently, fatal incidents involving highly dense crowds are so sudden that traditional methods of crowd-control are ineffective. This paper proposes the use self-forming “ad-hoc” P2P networks to detect crowd densities and using the sensors in the same device to perform “edge analytics” to detect anomalous crowd behavior. Autonomous crowd control can then be performed by transmitting warnings to the oncoming crowd via the same P2P network and further to a WAN if found.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 21% of the total text.

Real-time autonomous crowd control system via detection of crowd anomalies from peer-to-peer devices

By Kuan Heng, Lee, Colin Chee, Muhammad Umer Farooq,

Ondy Sukma, Oo Ewe Aik

Motorola Solutions

 

ABSTRACT

Recently, fatal incidents involving highly dense crowds are so sudden that traditional methods of crowd-control are ineffective. This paper proposes the use self-forming “ad-hoc” P2P networks to detect crowd densities and using the sensors in the same device to perform “edge analytics” to detect anomalous crowd behavior. Autonomous crowd control can then be performed by transmitting warnings to the oncoming crowd via the same P2P network and further to a WAN if found.

PROBLEM

Abnormal crowd behaviour detection and control are critical in very highly dense crowds. Poorly managed, it can result in many fatalities. Existing methods of crowd detection usually involve images from video feeds (e.g. video surveillance cameras) which are then post-processed and anomalies in crowd behaviour are usually detected too late. Some have proposed the use of personal sensors (e.g. cell phones), in areas where cameras do not exist and transmit positioning to nearby servers/cell towers. However, such systems get jammed in highly congested areas. Furthermore, even if the situation could be detected, the time needed for image processing is long and the issue of crowd control could be too late.

SOLUTION

Here we propose a real-time autonomous crowd control system that uses personal smart devices (e.g. smart phones/watches) with the ability to form a wireless ad-hoc peer-to-peer (P2P) network with the surrounding peers, to calculate the localized crowd anomalies through crowd density and motion pattern recognition. Based on these calculated information, the system will assess the risk of stampede and automatically providing warnings and instructions to the incoming peers via P2P messaging.

The main differentiation of our proposed system compared to the prior-art is that it does not require the mobile network connection and centralized server to detect anomalous crowd motion patterns and also provide real-time crowd control.

OPERATION

Figure 1: System topology of the invented crowd-control

As shown in Figure 1, the overall operation of the invented crowd control system can be divided into three levels. On the lowest level, the users install and run the crowd control application on their mobile devices, such as but not limited to laptops, smart phones, tablets, PDAs, MP3 and etc, which are powered by mobile OS such as but not limited to Android, iOS and Windows Mobile. With the permission of the user, the crowd control application will then enable the low power wireless connectivity (Bluetooth or WiFi) implemented on the respective mobile devices to scan for concurrent adjacent devices with similar crowd control application running. Subsequently, based on the scanned information, the crowd control application of the new member device will request to establish...