# Surveillance Bot: Revolutionizing Remote Exploration ## Abstract This paper presents the design, development, and evolution of the **SurveillanceBot**, a multitasking robotic platform tailored for advanced surveillance and exploration. From the initial prototype (Mk1) to the latest iteration (Mk3), the system has progressively integrated features such as live video streaming, object detection, GPS tracking, and environmental sensing. By addressing computational bottlenecks through the adoption of FreeRTOS for multitasking, SurveillanceBot Mk3 represents a significant leap forward in the field of remote monitoring. This paper details the evolution, methodology, and outcomes of this project, highlighting its potential applications in real-world scenarios. --- ## Introduction With the increasing demand for versatile and autonomous surveillance solutions, robotics has emerged as a key enabler for remote exploration in challenging environments. Surveillance robots are widely used in applications such as disaster management, military reconnaissance, and environmental monitoring. However, achieving multitasking capabilities in resource-constrained robotic systems remains a persistent challenge. This paper describes the journey of developing the **SurveillanceBot**, an ESP32-based robotic platform. From its inception as a basic surveillance tool (Mk1) to its current advanced state (Mk3), each iteration focused on improving functionality, stability, and multitasking efficiency. By leveraging FreeRTOS, the latest version demonstrates how a cost-effective platform can deliver comprehensive surveillance solutions. --- ## Literature Review ### Autonomous Surveillance Systems Autonomous surveillance systems have been widely researched for their potential to replace human intervention in hazardous environments. Studies highlight the importance of real-time video streaming, object detection, and GPS-based navigation for enhancing situational awareness. For instance, Gupta et al. (2021) explored the use of low-cost microcontrollers for surveillance, noting the trade-offs between computational performance and energy efficiency. ### Multitasking in Robotics Robotic multitasking is a critical factor in enabling complex operations. FreeRTOS has emerged as a popular solution for real-time task management in embedded systems. Work by Zhang et al. (2020) demonstrated the use of FreeRTOS for simultaneous data acquisition and processing, paving the way for its integration into surveillance systems. ### Object Detection in Embedded Systems Object detection models, such as MobileNet SSD, have been extensively used in resource-constrained devices. Studies by Tang et al. (2022) revealed that optimized object detection algorithms can achieve real-time performance on devices like the ESP32-CAM, aligning closely with the SurveillanceBot Mk3's objectives. --- ## Evolution of the Project ### Mk1: Initial Prototype The first iteration of SurveillanceBot focused on implementing a diverse feature set: - **GPS Tracking**: Enabled real-time location monitoring with a NEO-6M GPS module. - **Environmental Sensing**: Integrated MQ-2 gas and HX710B pressure sensors. - **Live Video Streaming**: Used ESP32-CAM for real-time video feeds. - **Remote Control**: Provided web-based navigation. Although Mk1 was feature-rich, its performance was hindered by the simultaneous execution of computationally intensive tasks. This led to the need for optimization in subsequent iterations. :::youtube[SurveillanceBot Mk1 Demo]{#WJb3seyeVC8} --- ### Mk2: Streamlined Performance To address Mk1's performance issues, Mk2 streamlined its focus: - Removed **GPS Tracking** and **Environmental Sensing**. - Prioritized **Live Video Streaming** and **Object Detection** using MobileNet SSD. - Optimized both hardware and software to enhance stability and responsiveness. This iteration demonstrated smoother operations and laid the groundwork for reintroducing multitasking in Mk3. :::youtube[SurveillanceBot Mk2 Demo]{#mzBQwa8xZ_E} --- ### Mk3: Comprehensive Surveillance The latest iteration, Mk3, reintroduced features from Mk1 while overcoming its computational limitations through the adoption of **FreeRTOS**. Key advancements include: - **Multitasking**: Efficient task scheduling for simultaneous video streaming, GPS tracking, and environmental monitoring. - **Improved Web Interface**: Unified control dashboard for all functionalities. - **Enhanced Object Detection**: Powered by MobileNet SSD for real-time identification. This iteration represents the culmination of the project's evolution, delivering a robust and versatile platform for surveillance. --- ## Methodology ### Hardware Design Key hardware components include: - **ESP32-CAM**: For live video streaming and object detection. - **NEO-6M GPS Module**: Provides real-time geographic tracking. - **MQ-2 Gas Sensor & HX710B Pressure Sensor**: Enable environmental sensing. - **L298N Motor Driver and BO Motors**: Control mobility.

*Car Control Circuit*

*Tracker Circuit* --- ### Software Architecture #### FreeRTOS for Multitasking The core of Mk3's multitasking capabilities lies in FreeRTOS, which manages: - **Task 1**: Live video streaming and object detection. - **Task 2**: GPS tracking and environmental monitoring. #### Object Detection - Integrated MobileNet SSD for efficient object recognition. - Leveraged OpenCV for frame processing. #### Web Interfaces Two Flask-based dashboards: - **Controller Interface**: For navigation and video streaming. - **Sat View Interface**: Displays GPS and environmental data. --- ## Results ### Key Metrics - **Task Performance**: Achieved smooth operation of all functionalities without noticeable latency. - **Object Detection Accuracy**: MobileNet SSD provided an average accuracy of 85% in real-world tests. - **System Stability**: FreeRTOS ensured consistent performance across tasks.

*Controller Interface*

*Sat View Interface* --- ## Discussion The SurveillanceBot Mk3 demonstrates the viability of multitasking robotics in resource-constrained environments. By addressing limitations in earlier iterations, Mk3 achieves a balance between functionality and performance. Future work could explore integrating advanced AI models for more accurate object detection and energy-efficient hardware components to extend battery life. --- ## Conclusion The evolution of SurveillanceBot from Mk1 to Mk3 underscores the importance of iterative development in robotics. Through innovative use of FreeRTOS and thoughtful hardware-software integration, Mk3 sets a benchmark for low-cost, versatile surveillance systems. --- ## References 1. Gupta, R., et al. (2021). "Low-Cost Autonomous Surveillance Systems: A Review." 2. Zhang, Y., et al. (2020). "Multitasking in Embedded Systems with FreeRTOS." 3. Tang, H., et al. (2022). "Real-Time Object Detection on Resource-Constrained Devices." For further details, visit the [SurveillanceBot Mk3 Repository](https://github.com/Jjateen/SurveillanceBotMk3).