<![CDATA[Abstract: Infantry operations in urban warfare or confined areas present high risks to personnel due to blind spots and ambush threats. The deployment of commercial military robots to address this issue is often hindered by prohibitive costs and low portability. This study aims to design and build a cost-effective and disposable tactical reconnaissance robot based on the ESP32-CAM microcontroller with a tracked chassis mechanism to support maneuvers in rubble terrain. The system is developed using a Client-Server architecture to transmit video and telemetry data, integrating image processing algorithms on the edge device for real-time human threat detection. Field tests were conducted to validate system performance in terms of communication range, latency, and detection accuracy. The results show that the robot operates effectively within a safe radius of 50-60 meters (Line of Sight) with an average video transmission latency of 160ms at HVGA resolution (320x240 pixels), which serves as the optimal point for control responsiveness. The object detection feature recognizes human faces with 95% accuracy at a distance of 1 meter, making it relevant as a close-range early warning system (Close Quarter Battle). This study demonstrates the feasibility of implementing low-cost technology as a strategic support asset to enhance the Situational Awareness (SA) UGV of infantry forces. Keywords: ESP32-CAM, Image Processing, Infantry, Reconnaissance Robot, Situational Awareness, UGV.]]>
