CCTV video surveillance systems are commonly used by businesses for property security. Requiring no changes to existing video surveillance infrastructure, Bigmate has developed an edge AI video analysis solution called Warny that enables AI computer vision capability in existing video surveillance systems. With Warny’s AI vision features, a high proportion of occupational accidents can be prevented, including those caused by behavioral mistakes, vehicle collisions and hazardous zone intrusions.
Employees are the most important assets of enterprises. Occupational accidents come at a significant cost to not only employees, but also the enterprises, society and economy.
During 2019 to 2020, according to Safe Work Australia, there were 120,355 serious work-related accidents in Australia which resulted in injury and disease to workers with absence from work of at least one working week. The compensation and lost time for each accident was on average $13,500 and 6.6 weeks.
Among serious work-related accidents, 23% of employees fell because of the work environment, 16% were hit by moving objects, and 7% hit objects with a part of their body. To prevent these types of accidents and improve workplace safety, ADLINK’s Australian AI solution partner, Bigmate, developed an edge AI video safety solution basing on the ADLINK DLAP-4000 edge AI platform to enable AI computer vision capability on existing CCTV video surveillance systems.
Bigmate is a business outcome focused technology with a mission to save lives and prevent workplace injuries. With over 20 years of experience, Bigmate has been developing market-leading AI solutions that leverage computer vision and machine learning to monitor the safety people working around fixed and mobile equipment.
Edge AI Video Analysis Solution for Workplace Safety
Bigmate leverages ADLINK’s DLAP-4000 edge AI platform for its Warny edge AI video analysis solution. Warny ingests existing CCTV video streams and analyses each frame using machine learning algorithms. Without changing the existing infrastructure, all that's required is simply connecting the edge AI video analysis platform to the DVR/NVR via an Ethernet cable. Powered by cutting-edge NVIDIA professional graphics solutions, the edge AI video analysis platform can analyse CCTV video streams at the edge, identifying incidents immediately and alerting stakeholders before accidents happen. Using Amazon Web Services (AWS), Warny can be remotely managed, presenting analysed data in dashboards. AWS also enables continuous training of AI models to near 100% accuracy.
DLAP-4000 Edge AI Video Analysis Platform
Powered by Intel® CPU and NVIDIA® GPU Technologies
Integrated Power Supply
Dual Gigabit Network Ports
Front Accessible I/O Ports
Supports NVIDIA Graphics Cards
We tested many available edge AI solutions from top-tier hardware providers including ADLINK. The DLAP-4000 stood out from a variety of edge devices. It met our minimum resource requirements with low power consumption and was compact in size. With ADLINK’s long life support and reputation in the industry, the DLAP-4000 is the best fit in all categories.
Sam Montague, APJ Partner Manager, Bigmate
Powerful processing performance supporting multiple video streams
Equipped with a 9th Gen Intel® Core™ i7 processor and scalable NVIDIA PCI Express graphics card, the DLAP-4000 can real-time analyse multiple video streams from CCTV systems. For applications that don’t require real-time alerts or machine integrations, the DLAP-4000 can process even more video streams depending on the use case.
Supports dual-width FHFL graphics card for accelerated AI inferencing
Featuring a PCI Express x16 Gen3 slot, the video processing capability of DLAP-4000 can swiftly be scaled up by adding an NVIDIA PCI Express graphics card to accomplish a diverse range of AI computer vision applications. In addition, ADLINK offers a comprehensive line-up of NVIDIA graphics cards, which benefits AI vision solution suppliers with a one-stop-shop for a broad range of graphics configurations.
Dual GbE ports for low-latency edge AI inferencing and continuous AI model training in the cloud
With the two GbE ports, the DLAP-4000 allows AI inferencing at the edge and optimizing AI models in the cloud at the same time. The DLAP-4000 is directly connected to DVR/NVR through one GbE port to the local network of the monitoring field site with low-latency to prevent accidents. While the DLAP-4000 is analysing video streams from surveillance systems, the DLAP-4000 is synchronizing captured events and optimized AI models with Amazon Web Services via the other GbE port. For use cases requiring machine integrations, the DLAP-4000 can also be connected to I/O modules via Ethernet to stop industrial machinery using MQTT.
Compact yet easy-to-install design with efficient power consumption, heat dissipation and front accessible I/O ports
In most warehouse and manufacturing occasions, CCTV systems were deployed many years before computer vision technology had matured. There was no pre-planned space for installing an additional edge AI video analysis platform to the DVR/NVR system. The DLAP-4000 occupies only 9.9 liters with a built-in power supply unit supporting up to 500W, enabling easy deployment in a limited space. With ADLINK’s sophisticated heat dissipation design experience, the DLAP-4000 can operate from 0°C to 50°C while supporting a PCI Express graphic card. The front accessible I/O ports ease installation and maintenance in space-limited racks.
ADLINK’s Edge AI Platform & PCI Express Graphics Cards
|8th/9th Gen Intel® Core™ i7/i5/i3 Processors
|Intel® H310 Chipset
|DDR4 2666/2400MHz, 2x SO-DIMM, up to 32GB
|PEG Card Support
NVIDIA RTX A2000
NVIDIA RTX A4000
NVIDIA RTX A5000
NVIDIA RTX A6000
|1x DVI-D, 1x DisplayPort, 1x HDMI
|4x USB 3.1,
2x USB 2.0
|1x PCIe x16 slot
Dual-Width Full-Height Full-Length
|2x 2.5" SATA 6Gb/s drive bays
|100 to 240 VAC
500W built-in PSU
|0°C to 50°C
(W x D x H)
|220 x 300 x 150mm
AI Vision-enforced Workplace Safety in Verticals
Warehousing & Logistics
- Vehicle collision prevention
- Identity recognition
Mining & Construction
- Personal protective equipment monitoring
- Restricted area monitoring
- Hazard prevention in human-robot collaboration
- SOP compliance reinforcement
Partnership between NVIDIA and ADLINK
ADLINK is a leading edge computing company with access to NVIDIA’s embedded GPU roadmap.
As an NVIDIA Jetson Preferred Partner, ADLINK gets direct support and early access from NVIDIA.