Design and Development of a Self-Balancing Patrolling Bike for Urban Surveillance and Mobility Enhancement

Abstract

As robotics continues to advance across industries, autonomous systems are increasingly being adopted for safety, surveillance, and human interaction. Urban environments, which are often large, complex, and difficult to monitor, still rely heavily on static cameras, manual patrols, and limited human oversight for security. These traditional methods become inadequate during peak periods such as night shifts and holidays, and they lack real-time updates, mobility, and multi-threat adaptability. This project proposes an autonomous patrolling bike designed to function as a mobile guard and digital sentry, capable of independently navigating predefined paths, avoiding obstacles, and providing real-time alerts and threat information. The robot integrates multimodal interaction through cloud connectivity, AI vision, and sensor fusion, supported by IoT infrastructure to access and update information dynamically. By combining mobility with intelligent communication, the system aims to enhance site safety, reduce risk for human guards, and demonstrate the effective role of robotics in creating smarter and more secure urban environments.

Introduction

The text presents the design and development of a self-balancing autonomous patrolling bike intended to enhance urban security in modern smart-city environments. Rapid urbanization, rising security demands, and the limitations of traditional surveillance methods—such as human fatigue, high operational costs, static camera blind spots, and limited mobility—have created a need for agile, automated security solutions capable of continuous operation.

The proposed system is a compact two-wheeled, inverted-pendulum robot that uses an ESP32 microcontroller and an MPU6050 IMU sensor to maintain dynamic balance. Its narrow structure allows it to navigate crowded and confined spaces inaccessible to conventional four-wheeled security vehicles. By integrating mobility with surveillance, the platform addresses coverage gaps caused by static CCTV systems and reduces reliance on human guards.

The patrolling bike is equipped with a 360-degree camera, obstacle-detection sensors, and cloud-based AI for advanced threat detection, including fire identification, crowd analysis, and anomaly detection. Heavy computational tasks are offloaded to cloud servers, enabling high performance while maintaining low cost and energy efficiency. Real-time alerts, GPS coordinates, and visual evidence are transmitted to authorities via platforms such as Telegram, significantly reducing response times during emergencies.

The text reviews existing security systems and highlights key challenges, including human limitations, environmental constraints, high costs of commercial robots, and insufficient integration between mobility, stabilization, and intelligent threat detection. A literature review demonstrates progress in self-balancing control, sensor fusion, machine vision, and IoT-based surveillance, while identifying gaps in affordable, agile, and fully integrated mobile security platforms.

To address these gaps, the project follows a structured mechatronic methodology involving modular hardware design, PID-based stabilization, autonomous navigation, cloud connectivity, and AI-driven detection. The system workflow ensures continuous balance through real-time control loops while simultaneously performing high-level security analysis in the cloud.

Conclusion

The design and implementation of the Self-Balancing Patrolling Bike demonstrate that autonomous service robots can significantly enhance urban safety by providing a mobile, intelligent, and cost-effective surveillance solution. By integrating high-frequency PID control for stabilization with cloud-based AI for threat detection, the project successfully bridged the gap between mechanical agility and advanced environmental intelligence. The system proved capable of operating independently, delivering consistent security information and real-time alerts without the need for constant human supervision. Testing and validation confirmed that the robot\'s modular architecture—comprising the ESP32 microcontroller, MPU6050 sensors, and YOLO-based vision—enables reliable obstacle avoidance and accurate hazard identification. The ability to maintain stability within $\\pm5^{\\circa}$ during motion and deliver critical alerts via Telegram within 5 seconds highlights the system\'s practical readiness for real-world deployment. Furthermore, the dual-mode control via a web dashboard ensures that the platform remains adaptable to both predefined patrol routes and manual emergency interventions. Ultimately, this project highlights how the convergence of robotics, IoT, and AI can create efficient and user-friendly security solutions for environments such as residential complexes, industrial sites, and public spaces. The successful development of a 1.5-foot, two-wheeled platform provides a versatile foundation for future advancements in autonomous patrolling, offering a scalable alternative to traditional static surveillance. Moving forward, the methodology established here can be expanded to include more diverse terrains and higher-complexity threat analysis, further contributing to the creation of smarter and more secure urban ecosystems.

References

[1] F. Grasser et al., “JOE: A Mobile, Inverted Pendulum,” IEEE Trans. Ind. Electron., vol. 49, no. 1, pp. 107–114, 2002. doi: 10.1109/41.982254. [2] Y. Kim et al., “Dynamic Analysis of a Nonholonomic Two-Wheeled Inverted Pendulum Robot,” Journal of Intelligent & Robotic Systems, vol. 44, pp. 25–46, 2015. doi: 10.1007/s10846-005-9022-4. [3] S. W. Nawawi et al., “Real-Time Control of a Two-Wheeled Inverted Pendulum Mobile Robot,” World Academy of Science, Engineering and Technology, vol. 29, pp. 214–220, 2008. [4] C. Huang et al., “Implementation of a Two-Wheeled Self-Balancing Robot by Using Kalman Filter and PID Controller,” Int. Conf. Fuzzy Theory and Applications, pp. 261–266, 2012. doi: 10.1109/iFuzzy.2012.6409740. [5] H. Li and L. E. Parker, “Cooperative Navigation and Surveillance in Distributed Multi-Robot Systems,” Proc. IEEE ICRA, pp. 1993–1998, 2008. doi: 10.1109/ROBOT.2008.4543501. [6] J. Redmon et al., “You Only Look Once: Unified, Real-Time Object Detection,” Proc. IEEE CVPR, pp. 779–788, 2016. doi: 10.1109/CVPR.2016.91. [7] S. Khan et al., “Energy-Efficient Deep CNN for Smoke Detection in Foggy IoT Environment,” IEEE IoT Journal, vol. 7, no. 9, pp. 8237–8248, 2020. doi: 10.1109/JIOT.2020.2986162. [8] A. Jain et al., “IoT Based Smart Surveillance Robot with Real Time Monitoring,” IEEE Int. Conf. on Power Energy, Environment and Intelligent Control, 2018. doi: 10.1109/PEEIC.2018.8665649. [9] M. Faisal et al., “Fuzzy Logic Control for a Self-Balancing Two-Wheeled Robot,” IEEE Int. Conf. on Control System, Computing and Engineering, pp. 285–290, 2013. doi: 10.1109/ICCSCE.2013.6719974. [10] S. Hussain et al., “Sensors and control strategies in autonomous mobile robotics,” Robotics and Autonomous Systems, vol. 95, pp. 1–12, 2017. doi: 10.1016/j.robot.2017.05.006. [11] R. Olfati-Saber, “Normal Forms for Nonholonomic Systems with Applications to Control,” IEEE Trans. on Automatic Control, vol. 47, no. 2, pp. 305–308, 2002. doi: 10.1109/9.983365. [12] K. Pathak et al., “Velocity and Position Control of a Self-Balancing Two-Wheeled Wheelchair,” IEEE Trans. on Robotics, vol. 21, no. 3, pp. 505–514, 2005. doi: 10.1109/TRO.2004.839237. [13] C. Li et al., “PID control for balancing robotic systems,” IEEE Trans. Industrial Electronics, vol. 61, no. 7, pp. 3770–3780, 2014. doi: 10.1109/TIE.2013.2270227. [14] A. Gupta and N. Sharma, “Cloud-connected surveillance using ESP32 and IoT,” IEEE Int. Conf. Computing, Power and Communication Technologies, 2018. doi: 10.1109/ICCPCT.2018.8453380. [15] J. S. Noh et al., “A Fault-Tolerant Control of a Two-Wheeled Inverted Pendulum Robot,” International Journal of Control, Automation and Systems, vol. 8, no. 4, pp. 823–832, 2010. doi: 10.1007/s12555-010-0415-3. [16] S. J. Anderson et al., “Autonomous navigation for urban patrolling robots,” IEEE Trans. on Intelligent Transportation Systems, vol. 11, no. 1, pp. 154–165, 2010. doi: 10.1109/TITS.2009.2033454. [17] S. Wang et al., “IoT-based real-time security monitoring,” IEEE Internet of Things Journal, vol. 6, no. 2, pp. 365–378, 2019. doi: 10.1109/JIOT.2018.2878687. [18] T. Nomura et al., “A Real-time Obstacle Avoidance for a Two-Wheeled Inverted Pendulum Robot,” Proc. IEEE/RSJ Int. Conf. on Intelligent Robots and Systems, 2009. doi: 10.1109/IROS.2009.5354467. [19] G. Chen et al., “Sensor fusion and IMU calibration for mobile robots,” Sensors, vol. 18, no. 9, p. 3090, 2018. doi: 10.3390/s18093090. [20] M. K. Zadeh et al., “IoT-integrated surveillance for smart city applications,” IEEE Sensors Journal, vol. 19, no. 22, pp. 10489–10502, 2019. doi: 10.1109/JSEN.2019.2931561.

Copyright

Copyright © 2026 Sharvani G R, Manjunatha K N, Druva R, Musaif Khan K, Chaya Kabadagi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.