--- lang: en title: "AI vision for retail, malls and restaurants" url: https://nabrio.com/industries/retail canonical: https://nabrio.com/industries/retail updated: 2026-10-27 publisher: Nabrio Co., Ltd. (nabrio.com) --- # AI vision for retail, malls and restaurants Turn existing cameras in malls, stores and restaurants into one system for visitor analytics, operations and security, processed on-site with privacy in mind. **Who it’s for:** Mall and property managers, retail operations and marketing heads, loss-prevention and security managers, quick-service restaurant operators, and the telecom and system-integrator partners who serve them. ## Why Nabrio in retail - **One system, three jobs:** Analytics, operations and security come from the same cameras and the same platform, instead of a separate vendor for each. - **Counts visitors correctly:** Person re-identification recognises the same visitor across entrances and devices, so nobody is counted twice. - **Processed on-site:** Video stays on your premises. Demographics are anonymous estimates, and face recognition is used only for watch-lists you define. - **Shares data with your systems:** Results flow to dashboards and to other systems over MQTT or HTTP. ## Customer stories - [120+ cameras and 20+ AI capabilities in one mall system](https://nabrio.com/customers/shopping-mall-ai-vision.md) (A shopping mall in Bangkok) ## Use cases ### Visitor analytics - [People counting in and out, with re-identification](https://nabrio.com/use-cases/people-counting.md): Simple door counters double-count people who return or use several entrances. - [Zone traffic and heatmaps](https://nabrio.com/use-cases/footfall-heatmaps-and-dwell-time.md): Cameras → tracking → visits per zone, paths, heatmaps → which areas, stores, and displays attract or lose traffic. - [Demographic analysis (age, gender)](https://nabrio.com/use-cases/visitor-demographics.md): Face or person crop → anonymous age and gender estimates → visitor profile by zone and time, to plan tenant mix and campaigns. - [Engagement analysis: dwell time and attention](https://nabrio.com/use-cases/footfall-heatmaps-and-dwell-time.md): Tracking gives the time spent in each zone; head pose and gaze show where people look (displays, windows, screens) → engagement per display or campaign. - [Expression and sentiment](https://nabrio.com/use-cases/visitor-demographics.md): Facial expression model (happy, sad, angry, surprised…) → mood trends by zone or event. - **Shopping-bag analytics**: Detect bags carried by visitors leaving → an indicator of conversion and purchases by entrance and time. - [Pet and pet-cart counting](https://nabrio.com/use-cases/people-counting.md): Detect pets and pet carts at entrances → counts for pet-friendly mall programmes. - [Event analytics](https://nabrio.com/use-cases/people-counting.md): The counting, zone traffic, demographics, and engagement above, applied to event spaces → attendance and audience profile per event, for organisers and sponsors. ### Store operations - [Staff presence and role monitoring](https://nabrio.com/use-cases/staff-presence-monitoring.md): Staffing must match demand and standards, but managers lack a real-time view across stores. - [Queue monitoring](https://nabrio.com/use-cases/queue-monitoring.md): Camera on the checkout or counter → count people in the queue zone and their waiting time → call more staff when the queue passes a threshold. - **Shelf and product availability**: Camera on shelves → empty-space detection or a "how full is this shelf?" score → restock alerts; planogram checks with a custom model. ### Security and loss prevention - [Loss prevention: asset removal](https://nabrio.com/use-cases/intrusion-detection.md): Watch high-value or fixed assets; alert when they are moved or removed - **Watch-list person detection**: Face recognition against an operator-defined watch-list, with anti-spoofing - [Intrusion detection](https://nabrio.com/use-cases/intrusion-detection.md): People in closed or restricted zones, after hours - [Height risk and climbing on railings](https://nabrio.com/use-cases/danger-zone-monitoring.md): Pose + zones at balconies and atriums - [Fire and smoke](https://nabrio.com/use-cases/fire-and-smoke-detection.md): Camera-based early warning - **Weapons (firearms, knives)**: Detect weapons and track the people involved across cameras - **Fights**: Violence and action recognition; track the people involved - [Falls and lying down](https://nabrio.com/use-cases/fall-detection.md): Pose action and fall detection; urgent alert - [Abandoned bags](https://nabrio.com/use-cases/intrusion-detection.md): Stationary unattended object in a zone longer than a set time ### Vehicles and parking - **Car traffic analytics**: Count vehicles in and out by hour; peak times - **Car park occupancy and dwell time**: Vehicles per bay or zone; how long each vehicle stays - [Licence plate recognition](https://nabrio.com/use-cases/licence-plate-recognition.md): Plate detection + OCR; VIP or tenant recognition; parking duration - **Car brand and model recognition**: Classify make and model, for customer profiling ## Connects to your systems Mall and store CCTV over RTSP, dashboards for operations and marketing teams, and partner or head-office systems over MQTT or HTTP. ## Frequently asked questions ### Does it identify shoppers? No. Age, gender and expression are anonymous estimates. Person re-identification works from overall appearance, not faces. Face recognition is used only for watch-lists the operator defines. ### Does it work with our existing CCTV? Usually, yes. Nara reads standard IP cameras over RTSP. Our shopping mall deployment runs on 120+ existing mall cameras. ### Where is the video processed? On-site, on your own hardware. Only the results you choose are shared with other systems.