ATAS Continuous Monitoring

AI Continuous Action Alert and Field Work Tracking using Vision and Sound

In areas where safety and security are vital, there is a need for continuous behaviour and scene surveillance using AI Vision, in a manner that is low cost and easy to implement. We now know human operators looking at CCTV cameras are expensive, prone to fatigue and distraction, leading to missed incidents and reduced vigilance over time. The subjective nature of human observation can also result in inconsistent and biased reporting, compromising the accuracy and reliability of workplace incident reporting. Additionally, manual surveillance is labour-intensive and costly, requiring continuous staffing and training, straining resources. These limitations highlight the need for more advanced, ultra secure, reliable, AI automated and cost scalable solutions to meet the growing demands of modern workplace security and safety.

Too many cameras, impossible to Watch them all.

Sites like Power, Oil and Gas Production require operators to be always alert

Remote Proctoring of Exam and Testing

Hotel Lobby Situational Awareness, how bad is the queue, terrorist threat etc.

Enter ATAS (AI Take Action System) Continuous Monitoring, an AI-powered surveillance systems for providing real-time monitoring and analysis, enhancing both safety and security. It is an innovative new system equipped with cutting-edge object detection abilities, fine-tuned to eliminate false positives. This advanced system can detect anomalies, ensure compliance with safety protocols, and prevent potential security breaches, all while minimizing human error. Our system is designed with a focus on being low cost and easy to use. It does not utilise any expensive API calls or require any special hardware, so it can even run on an office laptop or a small, embedded PC.

Key Features

• Report Generation: can record the incidents it observed and

collate them into reports for later viewing.

• No GPU Needed: The system can run in real time on CPUs of

existing PCs and notebooks.

• Alert without Cameras or Line of Sight: Exploits sound events

to signal strange situations like fighting, glass breaking.

• No Cameras Needed: No sophisticated cameras required, even a laptop webcam suffices for recording the environment.

• Offline Capable: Works without Internet, perfect for remote or

secure locations.

• High Wall Security: Backend system is military grade using

data diodes and AI security compliant.

• Low Cost: does not rely on expensive API calls, meaning the

whole system can be run cheaply and transparently. It is an instant

installation and you can start using immediately.

Use Cases

Construction Sites
Deploy at a construction site to monitor the environment and detect safety issues so as to protect the workers. It can monitor multiple sites at once and enforce safety standards more consistently than humans. It can also generate site safety reports to facilitate analysis of common issues and identifying workers who break the rules.

• PPE Detection: Detect if all workers are wearing the appropriate Personal Protective Equipment (PPE).
• Fatigue Detection: Detect if any worker looks tired and hence less alert and more prone to making mistakes.
• Equipment Malfunction: Detect unusual movements, smoke, leaks, and other signs of faulty equipment.
• Illegal Usage: Detect if unauthorised personnel attempt to access restricted areas or operate special equipment.
• Operator absence: Detect if personnel are missing from their stations when heavy machinery/equipment are in operation.

Data Centres
Deploy in a data centre to help enforce security and monitor for anomalies. This can greatly reduce the need for manual surveillance which is difficult in a large area.

• Intruder Detection: Detect when unauthorized

personnel are present in the data centre using

available servers by just adding webcam and microphone, even at the rack level.
• Tailgating Detection: Detect if any unauthorized individuals attempt to gain access to the data centre by closely following behind others.
• Access Control: Verify that authorized personnel are accessing only the appropriate areas of the data centre to complete their work.
• Equipment Malfunction: Detect potential issues with server racks, cooling systems or other infrastructure by identifying visual cues like smoke or leaks and listening for unusual sounds.