# Loitering Detection ::::{container} component_header_row
> [!NOTE] > This is a sample application **intended for evaluation and development purposes only**. > For more information, refer to > [Intended Use](https://docs.openedgeplatform.intel.com/2026.3/OEP-articles/notes-on-usage.html#intended-use) :::: Loitering Detection leverages advanced AI algorithms for monitoring and analyzing real-time video feeds to demonstrate the efficiency of Intel hardware in AI systems for detecting individuals lingering in specified areas and recognizing their activity. By utilizing cutting-edge technologies and pre-trained deep learning models, this application enables real-time processing and analysis of video streams, making it an ideal solution. Its modular architecture and integration capabilities ensure that users can easily customize and extend its functionalities to meet their specific needs. ## Key features - **Vision Analytics Pipeline:** Detect, track, and classify objects using pre-configured AI models, with polygon zone and dwell-time analytics computed natively in the pipeline (DL Streamer `gvaanalytics`). Customize parameters such as thresholds and zone polygons without requiring additional coding. - **Integration with MQTT and Grafana:** Facilitates efficient message handling, real-time monitoring, and insightful data visualization. - **User-Friendly:** Simplifies configuration and operation through prebuilt scripts and configuration files. - **Optional Console UI:** a single page to manage and compare different models simultanously. For details, see how to [Use the Console UI](./how-to-guides/use-console-ui.md). ## How it works The architecture is designed to facilitate seamless integration and operation of various components involved in AI-driven video analytics.  ### Components - **DL Streamer Pipeline Server (VA Pipeline):** Processes video frames, runs object detection and tracking, and computes zone presence and dwell time natively via the `gvaanalytics` element and a reusable `loitering_watermark` overlay element. - **Mosquitto Message Queuing Telemetry Transport (MQTT) Broker:** Facilitates message communication between the DL Streamer Pipeline Server and Grafana using the MQTT protocol. - **MQTT Table Flattener:** A small sidecar that reshapes the raw per-frame `object_tracking/