Industry application//Ruixun Technology AI-BOX in the coal m

Time：2023-07-04 Source：ROCKTECH Views：148

Introduction to this issue

The coal mining industry is one of the important material production sectors and pillar industries of my country's national economy. It plays an important role in improving living conditions, improving infrastructure, absorbing labor force employment, and promoting economic growth. At the same time, the coal mining industry is also a high-risk industry with frequent safety accidents. How to strengthen the safety management of the construction site, reduce the frequency of accidents, eliminate all kinds of illegal operations and uncivilized construction, and improve the quality of coal mine engineering will still be an important research topic in front of government departments at all levels, people in the industry and scholars.

Deeply integrated into the status quo of the core business of safety production, it is urgent to use information technology to carry out "intelligent" supervision of coal mine construction safety production. The remote management level of the site can speed up the speed of the enterprise to deal with the safety hazards of the engineering site.

As an Internet of Things company, Rocktech has been committed to connecting all things with technology, and we also have a presence in the coal mine industry. Today, we will share with you the application of human body recognition and alarm in mines:

Industry customers need to monitor the import and export of mines, intersections, working faces, ground and other major imports and exports, the statistics of the number of entrances and exits and personnel in different areas. Realize non-contact information collection and processing of long-distance moving targets in the tunnel, and timely prepare to reflect the dynamic situation of personnel in various areas in the mine tunnel to the central control room, so that managers can accurately grasp the distribution of personnel in the tunnel and the movement trajectory of miners at any time. In order to facilitate more reasonable scheduling and management of underground personnel, and when an accident occurs underground, the location of underground personnel can be found out in time for effective rescue.

According to this demand, Rocktech uses AI-BOX to communicate with cameras, alarm lights, alarm voices, routers and other peripherals to finally realize the capture and statistics of people entering and leaving specific areas; Snapshots and statistics, and modify specific areas through software as needed.

In view of the complex underground environment and the overall dark environment, workers will wear strong miner's lamps to enter. In this case, a network camera with high definition and adjustable focus is selected. AI-BOX communicates data with the back-end server through the router and network camera. , when someone breaks into the alarm area, the AI-BOX will take a snapshot and count in the system at the same time, upload the obtained data/pictures to the back-end server, and send instructions through the I/O port to light up while capturing and counting Alarm lights and alarm reminder equipment.

The red line in the figure is the counting trip line of the warning area. When someone enters/exits the alarm area from different directions, the system will automatically count the number of people entering/exiting, and at the same time, the alarm light and voice prompt function will be activated after entering. The direction and size of the line , The length can be adjusted according to demand. It can also realize that the warning area needs to be taken as the shooting range of the camera on site. Here, the warning area has been hidden, and at the same time, the functions of human body recognition, passing technology, snapshot of the warning area, and alarm reminder are realized.

In order to achieve the above performance, our big star AI-BOX builds a quad-core ARM Cortex-57 MPcore processor, configures 128 CUDA cores, uses LPDDR4 4GB memory, brings sufficient AI performance, provides 472GFLOP, and supports a series of AI frameworks and algorithms, 16GB eMMC storage, can run TENSORFLOW, PYTORCH, CAFFE/CAFFE2, KERAS, MXNET and other neural network, object detection, segmentation and language processing applications in parallel to realize image recognition, target detection and positioning, voice segmentation, video enhancement and intelligent analysis capabilities.

Using the latest video encoding H264/H265, it can process up to 5 high-definition full-motion video streams in real time, and can realize low-power edge intelligent video analysis performance for network video recorders, smart cameras and IoT gateways. In order to meet the full load of CPU and GPU, multiple cameras, and full display screen, YOLOv3 detects and recognizes objects, provides 4A power supply, and prevents extreme events such as device overheating and crashes.

AI-BOX

Technical Data Sheet

CPU	Quad-core ARM Cortex-A57 MPcore processor
GPU	Equipped with 128 CUDA cores
Memory	LPDDR4 4GB
Storage	16GB eMMC
video encoding	4K@30(H.264/H.265)
video decoding	4K@60(H.264/H.265)
Camera	12-lane MIPI CSI-2 DPHY 1.1(1.5Gbps
networking	Gigabit Ethernet port
display	HDMI/DP1.2
USB	4USB3.2 1Micro USB
Extension ports	1SDIO 2SPI 6I2C 2I2S
dimension	110.086.049.0mm

If you want to learn more about related solutions, you can leave your contact information and ask for detailed information.

ROCKTECH

Solutions

Industry application//Ruixun Technology AI-BOX in the coal m

Support

Products

Data download

Contact us