SalvAIoT: Intelligent AIoT System for Prevention and Support in Mountain Search and Rescue Operations
SalvAIoT is a multidisciplinary research project focused on developing an innovative AIoT (Artificial Intelligence of Things) system aimed at preventing mountain accidents and supporting Search and Rescue (SAR) operations. The system integrates wearable devices, fixed trail nodes, drones, LPWAN technologies (LoRa and NB-IoT), and TinyML algorithms to detect trail deviations and localize victims efficiently under real-world conditions with minimal energy and processing requirements.
Project Objectives
- Define the use cases, system architecture, and component specifications of the SalvAIoT platform.
- Design and develop mobile user devices (DMU) and trail marker devices (DMT) equipped with GPS, acoustic sensors, accelerometers, and LPWAN modules (LoRa/NB-IoT).
- Create a drone-mounted localization module (MLOC) for real-time victim detection using voice and LoRa/NB-IoT signals.
- Implement and validate TinyML algorithms for trail deviation detection and voice-based localization on embedded hardware.
- Integrate all hardware and software modules into the final SalvAIoT system and test in real mountain scenarios.
Research Focus
- Search and Rescue (SAR) technologies and mountain safety
- Low-power, wide-area communication systems for remote monitoring
- On-device AI using TinyML for audio/image recognition
- Resilient AIoT system design for extreme environments
Consortium and Partners
The project is coordinated by the Faculty of Electronics, Telecommunications and Information Technology, National University of Science and Technology Politehnica Bucharest (UPB)
Implementation Timeline
Start Date: October 09, 2023
End Date: December 31, 2025
Funding
Funded through a program of the National Alliance of Technical Universities (ARUT).
Total budget: €10,000
Expected Outcomes
- Fully integrated AIoT system for mountain accident prevention and SAR support
- TinyML models for deviation detection and victim voice localization
- Custom datasets with LoRa/NB-IoT signal measurements and voice samples
- Scientific publications, technical reports, and a potential patented solution
The SalvAIoT project aims to empower SAR teams and mountain trail users with intelligent, energy-efficient technologies that enhance safety, reduce response time, and promote responsible mountain activity through innovation in embedded AIoT systems.