Latency-Optimized Kinematic Control and Real-Time Visual Feedback for Edge-Intelligence Surveillance Platforms

Abstract

This study presents the design and implementation of a standalone mobile surveillance robot that enables real-time visual monitoring and remote navigation through wireless communication. The system is built around the ESP32-CAM Microcontroller Unit (MCU – Microcontroller Unit), which integrates image capture, processing, and communication within a compact embedded platform. The onboard camera captures visual data in Joint Photographic Experts Group (JPEG – Joint Photographic Experts Group) format at Video Graphics Array (VGA – Video Graphics Array) resolution, while efficient buffering is achieved using internal memory supported by Pseudo-Static Random Access Memory (PSRAM – Pseudo-Static Random Access Memory) to ensure continuous frame transmission. A persistent communication channel is established using the WebSocket protocol (WebSocket – FullDuplex Communication Protocol), enabling low-latency bidirectional exchange of control commands and video data. The robot’s movement is controlled using dual Direct Current motors (DC – Direct Current), interfaced through General Purpose Input Output pins (GPIO – General Purpose Input Output), with Pulse Width Modulation (PWM – Pulse Width Modulation) applied for precise speed regulation and smooth navigation. The software architecture follows an asynchronous event-driven model implemented using ESPAsyncWebServer and AsyncWebSocket libraries, allowing simultaneous execution of streaming and control operations without delay. The system operates via a Wireless Fidelity Access Point (Wi-Fi – Wireless Fidelity, AP – Access Point), enabling direct device connectivity without relying on external internet infrastructure. Additionally, a safety mechanism is incorporated to automatically stop the robot in case of connection loss, preventing unintended motion. Overall, the system provides a compact, efficient, and cost-effective solution for real-time surveillance and remote robotic control applications.