This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>

Search
  • Optical Fiber Sensing for Perimeter Inspection Solution

    Optical Fiber Sensing for Perimeter Inspection Solution

    Build a smart perimeter with a new generation of AI-powered optical fiber sensing.

  • Overview
  • Benefits
  • Architecture
  • Products
  • Case Studies
  • News and Events
  • Resources

Next-Generation AI Fiber Sensing Safeguards Perimeters

Intrusion incidents are common in large perimeters such as railways and airports, and in small campuses such as substations and oil and gas stations. These incidents can lead to significant losses in safe operations and property. Complex weather and environments — including vegetation, animals, pedestrians, vehicles, wind, rain, snow, and fog — put a great deal of pressure on the reliability, adaptability, and stability of perimeter protection systems. Perimeter protection — which traditionally requires largely manual efforts — is transforming into a more sophisticated type of technical protection. Due to its easy deployment and cost-effectiveness, the vibration optical cable is suitable for detecting various intrusions along long-distance fences. This is why it is regarded as one of the most important technical measures for protecting campuses.

Gain the Edge

Comprehensive detection, zero false negatives

Comprehensive detection, zero false negatives

The enhanced oDSP chip can detect DAS signals with a dynamic range of 50+ dB across the entire segment — four times higher than the industry average. This ensures that both strong and weak signals are collected without distortion. The chip parses DAS signals on the entire segment in real time, achieving μs-level latency and zero false negatives.

Anti-interference, low false positives

Anti-interference, low false positives

The multimodal sensing model analyzes intensity and phase data, as well as local short-term details and surrounding long-term features. It also generates alarms based on all-domain situation awareness while resisting interference from force-8 winds and heavy rains. It achieves a false positive rate smaller than 1 time/km/day.

Fast learning, continuous evolution

Fast learning, continuous evolution

The FDTS algorithm automatically optimizes sparse samples based on the FDTS gradient model and adapts to environment changes and user habits, improving the model library iteration efficiency 30-fold and continuously improving the alarm accuracy.

Architecture

Architecture

Architecture

(1) Optical cables are deployed on fences.

(2) The SAE sensing algorithm adapts to perimeter intrusions based on global environment features and eliminates interference from nature such as wind, rain, and small animals, implementing zero false negatives and anti-interference. It is also able to continuously evolve.

(3) The solution can flexibly work with the intelligent image identification feature of a video surveillance system to double check warnings, further improving user experiences.

Huawei

Real-World Success

TOP