Intelligent Swarm Robotics for Disaster Response and Obstacle Avoidance

Intelligent Swarm Robotics for Disaster Response and Obstacle Avoidance

연구 분야: Artificial Intelligence

학회: 2025 5th International Conference on Pervasive Computing and Social Networking (ICPCSN)

Swarm robotics relying on collective behaviour in biological systems like ant colonies, flocks of birds, and schooling fish poses a fault tolerant way to address hard problems under certainty in decentralized manners. Develop and function a swarm robotics system that comprises a multi-sensor and camera module commanded by an ESP32 microcontroller. The core capabilities are detection of humans and avoiding obstacles to perform efficient navigation through unfamiliar terrains. An internet-based website interface facilitates distant viewing and advanced planning through provision of real-time details in a comfortable format. A peer-to-peer low-latency communication provided by ESP-NOW creates a strong mesh network. LoRa integration allows the reliable longer-range transfer of data while retaining the low latency ESP-NOW peer-to-peer communication which extends the overall operational range. The combination of these layers guarantees robust inter-robot coordination, low power consumption, and reliable data transfer over various terrains. In the swarm mode, every ESP32 acts as an edge computing node processing light-weight machine learning models for onboard human detection. Eliminating centralized processing enables every robot to compute visual and sensor data locally and update the web server with information detected. This method is optimally suitable for search-and-rescue and autonomous operations.