Uavs & Drones

Anti-Drone System

Drone Communications Links

Davigations and Controls

Energies and Powers

Drone Task Load

Drone Launch & Recovery Systems

Drone Materials & Manufacturing

Drone Part Others

drone news today drone guide drone reviews drone supplier news drone show news drone news 2018
Here: Home > drone news > drone news today

Competition scheme: "5g + UAV" intelligent emergency rescue

  2019-12-03

introduce:China has a vast territory, complicated geology and natural conditions, and is a country with frequent natural disasters. In recent years, sudden and catastrophic events have frequently occurred, which seriously threatened people's lives and property

China has a vast territory, complicated geology and natural conditions, and is a country with frequent natural disasters. In recent years, sudden and catastrophic events have frequently occurred, which seriously threatened people's lives and property safety. At the same time, it also had a greater impact on social stability. Therefore, strengthening emergency management of emergencies is a major event that affects the overall economic and social development of the country and the safety of people's lives and property.

Facing the frequent occurrence of natural disasters and the severe and complex public safety situation, it is urgent to supplement the shortcomings of informationization, and comprehensively apply new technologies such as 5G networks, big data, the Internet of Things, cloud computing, and artificial intelligence, and build a system that is flat and flat. , Three-dimensional, intelligent, humane and other modern emergency rescue systems.

Against this background, the Sichuan Provincial Communications Administration, the Sichuan Provincial Emergency Management Department, and the Sichuan Federation of Trade unios decided to launch the "Sichuan 5G + Emergency Application Vocational Skills Contest" in May 2019, which is aimed at society, enterprises, and individuals. Extensively collect 5G + emergency applications, consolidate the capabilities and resources of the whole society, further explore the potential needs, business models, and technical support of 5G in emergency scenarios such as monitoring and early warning, safety production, fire prevention, and emergency dispatching, and deeply explore and incubate a batch of 5G + emergency features application.

Implementation time:

Project start-up: On May 17, 2019, the Sichuan Provincial Communications Administration, the Sichuan Provincial Emergency Management Department, and the Sichuan Provincial Federation of Trade unios jointly issued Sichuan Communications Bureau [2019] No. 2 — "about Holding 5G + Emergency Applications in Sichuan Province in 2019 Notice of vocational skills competition. "

Trial stage-before July 26, 2019, complete the system development and submit the entry plan. After researching and judging, Shulian decided to cooperate with its partners to use real-time stitching products based on drones to participate in the competition.

Finals preparation stage-On August 1, 2019, the number of entries in the number of joint programs finally stood out from 47 projects, and entered the finals with a good second place on the track.

The final stage-August 30, 2019, the number of joint solutions-"Zilian rescue-5G + drone intelligent emergency rescue solution" through the final defense, won the prize.


face the challenge

Industrial drones have outstanding advantages in applications such as air search and rescue and emergency rescue, such as high positioning accuracy, long battery life, fast emergency decision-making, and strong environmental adaptability. However, in the case of large-scale disaster rescue, there are certain shortcomings in the two mainstream drone emergency monitoring methods:

1. Although the drone equipped with a photoelectric pod can realize real-time live broadcast of the scene situation, it can only obtain local information and cannot provide commanders with the overall situation of the disaster scene;

2. The traditional aerial survey method of drones equipped with ortho cameras generally adopts the operation mode of “aerial photography-photo field control survey-photogrammetric aerial three encryption-model orientation-topographic mapping”. The mapping process is complicated and the production cycle is It is long, low in efficiency, and requires a large number of wild ground control points. For areas such as forests, mountains, and rivers where it is impossible to reach or find suitable ground control points, its application is limited, and it can not really take advantage of the rapid response and flexibility of uav remote sensing. The advantage of maneuverability restricts the application value of drones in emergency rescue.

In addition, the current interpretation of UAV images is mostly visual interpretation after the event, which cannot quickly and automatically extract disaster information. At the same time, the lack of fusion analysis of environmental data such as meteorology and geology is also not conducive to emergency commanders to carry out scientific and efficient Rescue decision.

To sum up, the problems of real-time processing of UAV images and intelligent decision analysis for landslides, debris flows, forest fires, floods and other large disasters need to be solved urgently. Currently, artificial intelligence-based drone image real-time processing and target interpretation technologies are still in their infancy. If successful, they will provide direct methods and technical support for the solution of this problem, and provide disaster prevention, mitigation and emergency response based on drones. Rescue research provides effective references, and is expected to broaden the application scenarios of industrial drones and provide important support for the development of industrial drones.

Schematic diagram of real-time video live broadcast with drone equipped with camera


Traditional UAV aerial photography and data processing flowchart

Application scenario

System advantages

"5G + UAV Intelligent Emergency Rescue System" has obvious advantages in three aspects: real-time, adaptability, versatility, and flexibility.

real-time

1. Real-time data transmission: based on the new-generation 5G cellular mobile communication network, it provides the networked drone with important capabilities such as real-time ultra-high-definition image transmission, remote low-latency control, and always online.

2. Real-time data processing: Using real-time image stitching and fast interpretation algorithm models based on technologies such as AI, computer vision, WebGIS, etc., the real-time returned image data can be processed in near real-time to achieve the entire decision-making process from the task. Real-time and automation.

Adaptability and versatility

1. Adaptability of UAV platform: The system adopts a new generation of pure electric vertical take-off and landing fixed-wing UAV platform, which can mount multiple types of mission loads, and has a long battery life (2-6 hours) and a high flight height (may be Up to 5000 meters), strong environmental adaptability (resistance to the wind of class 6) and other characteristics, has a strong adaptability.

2. The versatility of the software system: The input data of the real-time stitching system and intelligent analysis and decision-making system of drone images are only aerial images, image POS (Positioning and Orientation) data, and camera parameters. The data format has no special requirements, so the software system has no Versatility of human-machine platforms.

flexibility

1. UAV system flexibility: The system uses a new generation of pure electric vertical take-off and landing fixed-wing drone platform, which has the advantages of multi-rotor drone vertical take-off and landing and fixed-point surround, and the fixed-wing drone has fast speed and long flight time. It has the advantages of fast disassembly and assembly (completed within 1 minute), convenient transportation and carrying, etc., and can be put into use immediately in case of circumstances.

2. Software system deployment flexibility: Under different working modes, the two software systems of real-time stitching of UAV images and intelligent analysis and decision-making can be deployed in computer hardware as part of the ground system of the UAV, or they can be deployed in Data processing and analysis online in the cloud platform.

Application scenario

In summary, the system has the outstanding advantages of "global situational awareness, real-time data processing, and intelligent decision analysis", which can be widely applied to various application scenarios.

● Disaster emergency response: Emergency decision-making, command and rescue when earthquakes, landslides, mudslides, floods, forest fires and other large-scale disasters occur.

● Terrestrial surveying and mapping: For areas such as forests, mountains, and rivers that cannot be accessed or where suitable ground control points cannot be found, a rapid survey of the overall environment of the survey area is provided to provide a basis for route planning and control point layout.

● Public safety: meet the needs of large-scale conference environment patrol, emergency handling, border, sensitive area patrol and other needs.

● Target search: assist personnel search and rescue, crash search, aircraft, vehicle and other target search tasks.

● Environmental protection: Searching for straw incineration sites and searching for pollution sources.

data support

Due to the limited conditions at the competition demonstration site, it is inconvenient to use drones for on-site flight operations, so the live demo data used the drone monitoring data after the earthquake on June 20, 2019 in Shuanghe Town, Changning County (about 1.5km). UAV monitoring data mainly includes three types: aerial imagery, Position and Orientation System (POS), and camera parameters.

Aerial image

A total of 180 aerial photos with a resolution of 6000 * 4000 (3cm resolution), a total capacity of about 1.8G, and the file format is JPG.


POS data

The positioning and orientation system (POS) integrating dynamic GPS positioning technology and inertial navigation technology (INS) is applied to aerial remote sensing. It can directly obtain the six external azimuth elements of the image to achieve direct target positioning on the ground, breaking the traditional aerial photography. Surveying and GPS-assisted aerial photogrammetry must be limited by aerial triangulation.

The file format of POS data is CVS


Camera parameters

Due to differences in camera 0ion for each aerial photography operation, the system supports manual entry of camera parameters. The camera parameter template is shown below:


Application technology / implementation process

Technical solutions

The overall technical architecture of the "5G + Emergency Application Vocational Skills Competition in Sichuan Province in 2019" is shown below. The "5G + UAV Intelligent Emergency Rescue System" mainly belongs to the terminal perception layer and the comprehensive application layer, of which the drone search and monitoring system is the terminal Perception layer products, real-time stitching of UAV images, and intelligent analysis and decision-making systems belong to the comprehensive application layer.


5G + emergency application overall technical architecture

The overall business architecture of the "5G + UAV Intelligent Emergency Rescue System" is shown below.


5G + drone intelligent emergency rescue system overall business architecture

Operating mode

The working mode of "5G + UAV Intelligent Emergency Rescue System" mainly includes three types:

Working mode 1: local processing + 5G release

This mode is mainly aimed at the scenario where the basic network is severely damaged after a major natural disaster. It is integrated into the drone ground system by integrating software and hardware. During the drone's aerial flight, the aerial image and the aerial image are transmitted back in real time through the microwave image transmission link. POS data, real-time UAV image stitching and intelligent decision analysis and processing are performed locally, and the analyzed information is published on the Web through the 5G emergency communication network, so that the remote command center can quickly and real-time understand the situation on the scene.


Schematic of local processing + 5G release mode

Work mode 2: cloud processing + 5G network connected drone

In this mode, both the drone terminal and the ground control terminal transmit data and control instructions through the 5G network. The large bandwidth and low latency of the 5G network can realize real-time image transmission back to the cloud platform, and through the image deployed in the cloud Real-time stitching and intelligent decision analysis systems are used for business applications, integrating AI deep learning capabilities, fast element interpretation to achieve multi-method target locking and real-time monitoring.

However, this model relies more on the new air interface 5G antenna equipment of the drone. The current related technologies and equipment are still in the early stages of development, 5G chips and 5G terminals are not yet in place, and large-scale commercial applications are still immature.


5G network connected drone mode diagram

Work Mode 3: Cloud Processing + 5G Network Ground Station

In this mode, the image and control instruction transmission in the drone search system still uses the microwave communication link. The ground station sends local data to the cloud in real time through the 5G communication link (wired / wireless), and the images deployed in the cloud in real time. Splicing and intelligent decision analysis system for business applications.

In the early stage of the application of 5G technology, this mode is the best choice for 5G + UAV intelligent emergency applications, which is conducive to rapid promotion and implementation. This mode is also the actual working mode used in this competition.


5G Network UAV Ground Station Mode

System composition

"5G + UAV Intelligent Emergency Rescue System" is mainly composed of UAV search and monitoring system, real-time image stitching system and intelligent analysis and decision system. The following focuses on the "real-time stitching system for drone images" and "intelligent decision analysis system" developed by our company.

Drone image real-time stitching system

UAV image real-time stitching system is the nation's first set of fixed-wing unmanned aerial image real-time stitching software and hardware integrated products independently developed by our company, which can provide a one-stop solution real-time data collection to data processing and analysis at the emergency site. The solution realizes the real-time global perception of the situation on the scene, and achieves the effect of "seeing what can be seen and what can be seen". It helps users to solve the problems of separation of internal and external business and long response period in traditional aerial surveys, and comprehensively improves the efficiency of aerial survey operations, which can effectively respond to the needs of fast response and efficient decision-making in emergency scenarios.


Intelligent analysis decision system

The intelligent analysis and decision-making system can automatically extract the disaster target in the image, and quickly collect information such as quantity and area. At the same time, after integrating environmental factors such as meteorology and geology, it can assist experts in fast and scientific comprehensive command and dispatch.


External cooperation

External partners mainly include: Sichuan Aosi Technology Co., Ltd., China Telecom Co., Ltd. Sichuan Branch, Huawei Technologies Co., Ltd., China Tower Corporation Ziyang Branch, Chengdu New Orange Beidou Zhilian Co., Ltd.

1. Sichuan Aoshi Technology Co., Ltd. is responsible for the design and construction of "drone search and rescue system", which specifically includes four parts of the drone platform, ground control station, data link and multi-tasking load, which belongs to the perception layer in large systems. product.

2. China Telecom Co., Ltd. Sichuan Branch is responsible for providing the 5G coverage and network access required for the competition demonstration area and demonstration site, ensuring the data connection between the drone search and rescue system and the drone data processing cloud platform.

3. Huawei Technologies Co., Ltd. is responsible for providing Huawei's public cloud infrastructure platforms: network, computing, storage, security, virtual machines and other public cloud IaaS resources, and ensuring the smooth operation of the UAV image real-time stitching system and intelligent analysis and decision-making system.

Business change

"5G + UAV Intelligent Emergency Rescue Solution" will deeply integrate UAV and 5G technology, combined with the real-time stitching system of UAV images independently developed by Sulian, will deal with natural disasters, accident disasters and social security events. It plays an important role, provides effective reference for the research of disaster prevention and mitigation based on drones and emergency rescue, assists the development of China's aviation emergency rescue system, effectively promotes the transformation of the "5G + drone" disaster prevention emergency industry, and has a huge industry value.

The comparison between traditional emergency aerial survey methods and computer vision-based intelligent processing methods is shown in the following table:


Note: The data of "performance comparison" in the table is calculated based on the drone monitoring data after the earthquake in Shuanghe Town, Changning County on June 20, 2019.

Case submission companies · Lianlian:

The company started in 2012, led by Zhou Tao, a big data industry leader, and has more than 220 members. For many years, Shulian has been committed to becoming a big data industry ecological builder integrating data, algorithms and applications, and providing partners with a big data industry integrating top-level design, metadata management, analysis and mining, and business intelligence visualization. Chain services help companies to quickly transform in the wave of the digital economy and win market opportunities. Shulian has established long-term partnerships with a number of industry benchmarking companies, and obtained the lead investment of Dingxing Quantum in October 2018. Goldman Capital, Ginkgo Valley Capital, and China Science and Technology Innovation Fund followed the A + round of financing. After the completion of the round of financing, the value of the connected market has exceeded US $ 100 million.

Relate news

    Home | About us | Contact Us | sitemap | Rank | Advertising | gift | guestbook | RSS