Project background
1. The 21st century is the era of the Internet and AI. Using intelligent robots to replace patrol guards for repetitive and essential inspections is a future trend.
2. Substations are often outdoors. Human inspection has drawbacks: complex roads, low shrubs, flowerbeds, and sculptures can create security blind spots. Manual inspections can easily miss faults or record errors, especially during night patrols when fatigue increases the chance of missed routine checks.
3. For power equipment inspections, humans may not easily detect equipment anomalies or temperature abnormalities, and personnel are not highly sensitive to early fire signs.
4. Advantages of robot inspection: robots are less prone to fatigue, oversight, or human error. Patrol schedules, routes, and alarm schemes can be preconfigured to ensure timely and comprehensive coverage, strengthen scanning of blind spots, and use sensors, IoT, and artificial intelligence for real-time anomaly detection through image analysis.
Project goals
1. The robot includes infrared detection to identify high-temperature areas that are hard for people to notice and to provide early fire warnings. The robot transmits data reliably in dual-mode over 4G and Wi-Fi 6 to local or remote endpoints, enabling real-time monitoring and control from PCs or mobile devices.
2. The intelligent inspection robot provides long-range, omnidirectional, automatic photo capture, video recording, and remote monitoring for automated early warning. It can be extended to include temperature and humidity, smoke, and noise monitoring alarms, video image analysis, and danger pattern recognition. It maximizes mobile dynamic video capture, supports preset positions and patrol routes, and can integrate with centralized video management systems. System design follows applicable regulatory requirements.
3. Control protocols, video codec formats, interface protocols, video file formats, and transport protocols follow relevant national standards. Component selection and implementation consider overall system reliability.
4. This project uses the Renesas RA6 series as the main controller and accepts RTK positioning. The RTT-HMI-Board serves as the human-machine interaction control hub. The robot supports real-time image analysis with or without network connectivity. Face recognition algorithms run at the edge; edge AI provides lower latency, enhanced privacy protection, and improved reliability.
System architecture
GUI interface
Note: Wi-Fi or a local network connects to the robot via IP handshake. User login and access control are required; after configuration the robot begins inspection.
Development environment
RT-Thread Studio, LVGL, e2 studio, fsp_v4_4_0
LVGL GUI components include Label, Button, Image, Image Button, Keyboard, Calendar, and Chart.
