<![CDATA[Search and rescue (SAR) operations often require access to areas that are hazardous or unreachable for human rescuers. This study developed ArachnoSAR, a low-cost four-legged (quadruped) spider robot prototype that integrates legged locomotion, camera-based human detection, and automatic notification. The robot is driven by eight SG90 servo motors controlled by an ESP32, while a Raspberry Pi processes the camera stream using a Haar Cascade classifier and sends detection photographs to a Telegram bot, accompanied by a buzzer alert. Preliminary indoor testing showed that the system detected frontal faces at distances of up to 300 cm at 12 to 17 frames per second, while locomotion tests recorded speeds of 0.55 cm/s on ceramic and 1.36 cm/s on asphalt surfaces, with an estimated operating time of about 51 minutes. Identified limitations include the requirement for frontal face orientation, sensitivity to low lighting, motion blur, and dependence on internet connectivity for notification. These results indicate that the prototype is feasible as a proof of concept, while highlighting the need for image stabilization, deep learning based detection, and offline communication for real SAR deployment.]]>
Copyrights © 2026
