Automated Detection and Prioritization of Civic Infrastructure Complaints Using a Cascaded YOLOv8–LLM Pipeline with Geospatial Priority Scoring on Serverless AWS
Keywords:
Civic complaint management, Geospatial priority scoring, Large language models, Object detection, Serverless computing, Smart cities, YOLOv8Abstract
The systems used for lodging complaints by the municipalities in most Indian cities involve manual indenting, which leads to delays, inaccurate information, and wrong routing of complaints to departments. This paper presents a serverless web application that automates all the stages in the civic complaint life cycle, which prolongs the whole process. In this system, the complaint is submitted through images only, and the citizens do not need to define or classify the complaint. The important aspect of this system is that it works based on the seven-phase AI pipeline, which incorporates a specially trained YOLOv8 detector to determine 4 most common types of complaints: potholes, garbage dumps, waterlogging, and faulty streetlights. In case of low confidence prediction can always fall back on Amazon Nova Lite, a multimodal large language model that can be accessed using Amazon Bedrock. Furthermore, the pipeline also performs the assessment of the severity of the complaint, automatic routing to a department, generation of complaint descriptions, and priority scoring. The backend part of the application is executed by means of AWS Lambda, S3, API Gateway, and DynamoDB. The mentioned YOLOv8 has an mAP@50 indicator equal to 0.75 for all 4 classes.
References
United Nations, Department of Economic and Social Affairs, Population Division, World Urbanization Prospects: The 2018 Revision. New York, NY, USA: United Nations, 2019.
Ministry of Road Transport and Highways, Government of India, “Potholes Related Crashes on NHs,” Lok Sabha Unstarred Question No. 2175, answered Feb. 12, 2026, New Delhi, India.
T. Brown et al., “Language Models Are Few-Shot Learners,” in Advances in Neural Information Processing Systems, H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, Eds. Red Hook, NY, USA: Curran Associates, 2020, vol. 33, pp. 1877–1901.
J. Redmon, S. K. Divvala, R. B. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” in Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016, pp. 779–788.
D. Arya, H. Maeda, S. K. Ghosh, D. Toshniwal, and Y. Sekimoto, “RDD2022: A multi-national image dataset for automatic road damage detection,” Geosci. Data J., vol. 11, no. 4, pp. 846–862, 2024.
P. S. Vinu, T. I. V. Thanveerul, A. R. A. Abdur Rahim, and J. R. S. Jancy Rani, “AI-Driven Smart City Grievance Management and Automated Complaint Escalation System,” International Journal for Research in Applied Science & Engineering Technology, vol. 14, no. 4, pp. 1451–1455, Apr. 2026.
H. Maeda, Y. Sekimoto, T. Seto, T. Kashiyama, and H. Omata, “Road damage detection and classification using deep neural networks with smartphone images,” Computer-Aided Civil and Infrastructure Engineering, vol. 33, no. 12, pp. 1127–1141, 2018.
N. Bhavana, M. M. Kodabagi, B. M. Kumar, P. Ajay, N. Muthukumaran, and A. Ahilan, “POT-YOLO: Real-Time Road Potholes Detection Using Edge Segmentation-Based Yolo V8 Network,” IEEE Sensors J., vol. 24, no. 15, pp. 24802–24809, 2024.
Y. Ding et al., “Application of Improved YOLOv8 Image Model in Urban Manhole Cover Defect Management and Detection: Case Study,” Sensors, vol. 25, p. 4144, 2025.
M. Manu and A. G. L. Ajith, “Smart Complaint System Using Generative-AI,” Zhuzao/Foundry, vol. 28, no. 8, pp. 241–250, 2025.
S. Salunkhe, T. Salunke, A. Saware, A. Salvi, A. Patil, and R. Jolhe, “CivicMate: An AI-Driven City Grievances,” in 2026 6th Biennial International Conference on Nascent Technologies in Engineering (ICNTE), 2026, pp. 1–6.
V. Ishakian, V. Muthusamy, and A. Slominski, “Serving deep learning models in a serverless platform,” 2018 IEEE International Conference on Cloud Engineering (IC2E), Orlando, FL, USA, 2018, pp. 257–262.
S. Eismann et al., “Serverless applications: Why, when, and how?,” IEEE Software, vol. 38, no. 1, pp. 32–39, 2021.
G. Jocher, J. Qiu, and A. Chaurasia, Ultralytics YOLOv8, software, ver. 8.0.0, 2023.
M. Ester, H.-P. Kriegel, J. Sander, and X. Xu, “A density-based algorithm for discovering clusters in large spatial databases with noise,” in 2nd International Conference Knowledge Discovery and Data Mining (KDD), Portland, OR, USA, 1996, pp. 226–231.
S. Khanuja et al., “MuRIL: Multilingual Representations for Indian Languages,” arXiv:2103.10730 [cs.CL], 2021.
