Case Study · AI & Automation

PoolScout Camera, AI and mobile alerting for pool monitoring

Partner project by SoftTeco

Project profile

CustomerDeep Innovations (PoolScout)
PartnerSoftTeco
IndustrySmart home / IoT

Challenge

Deep Innovations wanted to make the existing PoolScout application more stable and more precise. Unpredictable behavior and false alarms made real-time monitoring difficult.

Solution

Improving how the components work together

SoftTeco analyzed the architecture and the data flows. The rework addressed concurrent access to shared data, among other things. A camera calibration accounts for pool boundaries and capture parameters.

Detecting and notifying

Computer vision distinguishes objects and behavior. Camera, alarm unit and mobile app work together; live video and configurable notifications support the response of the stored contacts.

Detecting, then informing along several paths

Camera at the pool: Live video → AI analysis: Objects / behavior; AI analysis: Objects / behavior → Configured rules: Event assessment; Configured rules: Event assessment → Alarm unit: Warning signal; Configured rules: Event assessment → Mobile app: Video / alert; Mobile app: Video / alert → Stored contacts: Notification.

Technologies

Flutter
Node.js / SQL
Computer vision / Wi-Fi

Results

According to SoftTeco, the modernization improved stability and monitoring quality and reduced false alarms. Specific error rates are not stated. The system is presented as support, not as a guarantee against accidents or a substitute for supervision.

More stable real-time monitoring
Fewer false alarms
Live video and notifications