A Comprehensive Survey and Reference Architecture for AI-Powered Autonomous Drone Systems in Smart Cities

Abstract

Despite a rapid rise of AI-powered Unmanned Aerial Vehicle (UAV) deployments in smart city environments, current surveys and frameworks lack a unified, protocol-level reference architecture that integrates multi-domain applications, edge AI perception, cognitive reasoning through Large Language Models (LLMs), and regulatory compliance within a single deployable specification. This study presents a comprehensive cross-domain review of AI-powered drone systems for traffic management, delivery, infrastructure inspection, disaster response, and environmental monitoring. The study introduces COMPASS (Cognitive Operations Model for Programmable Autonomous Smart-city Systems), a novel seven-layer technical reference architecture that describes communication protocols (MAVLink 2.0, ROS2/DDS, MQTT 5.0, and NGSI-LD), edge computing hardware recommendations for five drone payload tiers, and quantified performance requirements for safety-critical operations. The key feature of COMPASS is its LLM-based Semantic Middleware Layer, which allows for context-aware decision-making, natural human-drone interaction, and regulatory compliance verification. Comparing COMPASS to many other frameworks reveals that it is the only architecture to simultaneously provide multi-domain coverage, protocol-level specifications, hardware recommendations, LLM integration, and empirically verified benchmarks.

Keywords

Unmanned aerial vehicles / Artificial intelligence / Smart cities / Reference architecture / Edge computing

References

[1] Mohamed N, Al—Jaroodi J, Jawhar I, Idries A, Mohammed F. Unmanned aerial vehicles applications in future smart cities. Technol. Forecast. Soc. Change 2020, 153, 119293. DOI: 10.1016/j.techfore.2018.05.004

[2] Habibzadeh H, Soyata T, Kantarci B, Boukerche A, Kaptan C. Sensing, communication and security planes: A new challenge for a smart city system design. Comput. Netw. 2018, 144, 163-200. DOI: 10.1016/j.comnet.2018.08.001

[3] Hassanalian M, Abdelkefi A. Classifications, applications, and design challenges of drones: A review. Prog. Aerosp. Sci. 2017, 91, 99-131. DOI: 10.1016/j.paerosci.2017.04.003

[4] Důbravová H, Bureš V, Velfl L. Review of the application of drones for smart cities. IET Smart Cities 2024, 6, 312-332. DOI: 10.1049/smc2.12093

[5] Ezzat MA, Abd El Ghany MA, Almotairi S, Salem MA—M. Horizontal review on video surveillance for smart cities: Edge devices, applications, datasets, and future trends. Sensors 2021, 21, 3222. DOI: 10.3390/s21093222

[6] Hossain S, Lee D—J. Deep learning—based real—time multiple—object detection and tracking from aerial imagery via a flying robot with GPU—based embedded devices. Sensors 2019, 19, 3371. DOI: 10.3390/s19153371

[7] Lin T—Y, Goyal P, Girshick R, He K, Dollár P. Focal loss for dense object detection. In Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 22—29 October 2017; pp. 2980-2988.

[8] Hayat S, Yanmaz E, Muzaffar R. Survey on unmanned aerial vehicle networks for civil applications: A communications viewpoint. IEEE Commun. Surv. Tutor. 2016, 18, 2624-2661. DOI: 10.1109/COMST.2016.2560343

[9] Alhabib MH, Ali QI. Internet of autonomous vehicles communication infrastructure: A short review. Diagnostyka 2023, 24, 1-9. DOI: 10.29354/diag/168310

[10] Wei J, Wang X, Schuurmans D, Bosma M, Xia F, Chi E, et al. Chain—of—thought prompting elicits reasoning in large language models. Adv. Neural Inf. Process. Syst. 2022, 35, 24824-24837. DOI: 10.52202/068431—1800

[11] Bommasani R. On the opportunities and risks of foundation models. arXiv 2021, arXiv:2108.07258. DOI: 10.48550/arXiv.2108.07258

[12] Garlan D, Shaw M. An introduction to software architecture. Adv. Softw. Eng. Knowl. Eng. 1993, 1, 1-39. DOI: 10.1142/9789812798039_0001

[13] Shi W, Cao J, Zhang Q, Li Y, Xu L. Edge computing: Vision and challenges. IEEE Internet Things J. 2016, 3, 637-646. DOI: 10.1109/JIOT.2016.2579198

[14] Vemprala SH, Bonatti R, Bucker A, Kapoor A. Chatgpt for robotics: Design principles and model abilities. IEEE Access 2024, 12, 55682-55696. DOI: 10.1109/ACCESS.2024.3387941

[15] Kumar V. Autonomous Drones for Disaster Risk Reduction. AI for Good, International Telecommunication Union (ITU), 19 October 2022. Available online: https://aiforgood.itu.int/autonomous—drones—saving—lives—and—powering—disaster—preparedness/ (accessed 7 January 2026).

[16] Sze V, Chen Y—H, Yang T—J, Emer JS. Efficient processing of deep neural networks: A tutorial and survey. Proc. IEEE 2017, 105, 2295-2329. DOI: 10.1109/JPROC.2017.2761740

[17] Alsamhi SH, Ma O, Ansari MS, Almalki FA. Survey on collaborative smart drones and internet of things for improving smartness of smart cities. IEEE Access 2019, 7, 128125-128152. DOI: 10.1109/ACCESS.2019.2934998

[18] Nex F, Armenakis C, Cramer M, Cucci DA, Gerke M, Honkavaara E, et al. UAV in the advent of the twenties: Where we stand and what is next. ISPRS J. Photogramm. Remote Sens. 2022, 184, 215-242. DOI: 10.1016/j.isprsjprs.2021.12.006

[19] Javaid S, Khan MA, Fahim H, He B, Saeed N. Explainable AI and monocular vision for enhanced UAV navigation in smart cities: Prospects and challenges. Front. Sustain. Cities 2025, 7, 1561404. DOI: 10.3389/frsc.2025.1561404

[20] Garg T, Gupta S, Obaidat MS, Raj M. Drones as a service (DaaS) for 5G networks and blockchain—assisted IoT—based smart city infrastructure. Clust. Comput. 2024, 27, 8725-8788. DOI: 10.1007/s10586—024—04354—1

[21] Redmon J, Divvala S, Girshick R, Farhadi A. You only look once: Unified, real—time object detection. In Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 27—30 June 2016; pp. 779-788.

[22] Bochkovskiy A, Wang C—Y, Liao H—YM. Yolov4: Optimal speed and accuracy of object detection. arXiv 2020, arXiv:2004.10934. DOI: 10.48550/arXiv.2004.10934

[23] Boudjit K, Ramzan N. Human detection based on deep learning YOLO—v2 for real—time UAV applications. J. Exp. Theor. Artif. Intell. 2022, 34, 527-544. DOI: 10.1080/0952813X.2021.1907793

[24] Wang C, Han Y, Yang C, Wu M, Chen Z, Yun L, et al. CF—YOLO for small target detection in drone imagery based on YOLOv11 algorithm. Sci. Rep. 2025, 15, 16741. DOI: 10.1038/s41598—025—99634—0

[25] Zhang Z. Drone—YOLO: An efficient neural network method for target detection in drone images. Drones 2023, 7, 526. DOI: 10.3390/drones7080526

[26] Alshaer N, Abdelfatah R, Ismail T, Mahmoud H. Vision—Based UAV Detection and Tracking Using Deep Learning and Kalman Filter. Comput. Intell. 2025, 41, e70026. DOI: 10.1111/coin.70026

[27] Merza ME, Hussein SH, Ali QI. Identification scheme of false data injection attack based on deep learning algorithms for smart grids. Indones. J. Electr. Eng. Comput. Sci. 2023, 30, 219-228. DOI: 10.11591/ijeecs.v30.i1.pp219—228

[28] Rohan A, Rabah M, Kim S—H. Convolutional neural network—based real—time object detection and tracking for parrot AR drone 2. IEEE Access 2019, 7, 69575-69584. DOI: 10.1109/ACCESS.2019.2919332

[29] MAVLink. MAVLink Developer Guide. Available online: https://mavlink.io/en/ (accessed on 15 December 2025).

[30] Ibrahim Q, Qassab M. Theory, Concepts and Future of Self Organizing Networks (SON). Recent Adv. Comput. Sci. Commun. 2022, 15, 904-928. DOI: 10.2174/2666255814666210201102326

[31] Shah SFA, Mazhar T, Al Shloul T, Shahzad T, Hu Y—C, Mallek F, et al. Applications, challenges, and solutions of unmanned aerial vehicles in smart city using blockchain. PeerJ Comput. Sci. 2024, 10, e1776. DOI: 10.7717/peerj—cs.1776

[32] Ren S, He K, Girshick R, Sun J. Faster r—cnn: Towards real—time object detection with region proposal networks. Adv. Neural Inf. Process. Syst. 2015, 28. Available online: https://proceedings.neurips.cc/paper/2015/hash/14bfa6bb14875e45bba028a21ed38046—Abstract.html (accessed on 5 January 2026).

[33] al. GJe. Ultralytics YOLO. Available online: https://github.com/ultralytics/ultralytics (accessed on 17 December 2025).

[34] Ullah I, Adhikari D, Khan H, Ahmad S, Esposito C, Choi C. Optimizing mobile robot localization: Drones—enhanced sensor fusion with innovative wireless communication. In Proceedings of the IEEE INFOCOM 2024—IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), Vancouver, BC, Canada, 20 May 2024; pp. 1-6.

[35] Akhtar MT, Abe M, Kawamata M. A new variable step size LMS algorithm—based method for improved online secondary path modeling in active noise control systems. IEEE Trans. Audio Speech Lang. Process. 2006, 14, 720-726. DOI: 10.1109/TSA.2005.855829

[36] Premachandra C, Kunisada Y. GAN based audio noise suppression for victim detection at disaster sites with UAV. IEEE Trans. Serv. Comput. 2023, 17, 183-193. DOI: 10.1109/TSC.2023.3338488

[37] Luo Z, Ma H, Shi D, Gan W—S. Gfanc—rl: Reinforcement learning—based generative fixed—filter active noise control. Neural Netw. 2024, 180, 106687. DOI: 10.1016/j.neunet.2024.106687

[38] NVIDIA. Pie from the Sky: Drone Startup Delivers Pizza, Meds and Side of Excitement. NVIDIA Blog, 13 December 2023. Available online: https://blogs.nvidia.com/blog/zipline—drone—jetson—inception/ (accessed 17 December 2025).

[39] Khanna T, Gonzalez G. Zipline: Expanding the World’s Largest Autonomous Drone Delivery Network ; Harvard Business School: Boston, MA, USA, 2024.

[40] Google. Edge TPU inferencing overview. Coral, Google LLC. Available online: https://www.coral.ai/docs/edgetpu/inference/ (accessed on 17 December 2025).

[41] Intel. Intel Movidius Vision Processing Units ; Intel: Santa Clara, CA, USA, 2024.

[42] Betalo ML, Ullah I, Tesema FB, Wu Z, Li J, Bai X. Generative AI—Driven Multi—Agent DRL for task allocation in UAV—Assisted EMPD within 6G—Enabled SAGIN networks. IEEE Internet Things J. 2025, 12, 35890-35907. DOI: 10.1109/JIOT.2025.3579780

[43] Navot A, Wells S. How Amazon's delivery drones make smart landing decisions in unexpected conditions. About Amazon, Amazon.com, Inc., 2 September 2025. Available online: https://www.aboutamazon.com/news/transportation/prime—air—amazon—drone—delivery—faa—safety (accessed on 15 December 2025).

[44] Axios. 2024 Will Be a Breakout Year for Delivery Drones. Available online: https://www.axios.com/2024/01/02/delivery—drones—2024—amazon—zipline—wing (accessed on 15 December 2025).

[45] Bai X, Li B, Ullah I, Wu Z, Basheer S, Bashir AK. Energy—efficient routing for IoT—enabled multi—truck multi—drone pickup and delivery systems. Appl. Energy 2025, 400, 126546. DOI: 10.1016/j.apenergy.2025.126546

[46] Khanna T, Gonzalez G. Zipline: The World’s Largest Drone Delivery Network ; Harvard Business School: Boston, MA, USA, 2020.

[47] Manufacturing Today. Delivery Drones Take off as Wing and Amazon Expand; Manufacturing Today: Norfolk, UK, 2025.

[48] Aviation W. Wing Delivery Network. Available online: https://wing.com/ (accessed on 15 December 2025).

[49] Averroes. Utilizing Drones for Infrastructure Inspection. Available online: https://averroes.ai/blog/infrastructure—inspection—drones (accessed on 17 December 2025).

[50] FlytBase. Streamlining Infrastructure Inspections: The Power of AI and Docked Drones ; FlytBase: Pune, India, 2025.

[51] Odeh I, Giacco M, Islam T, Pawar KK, Ashar V. The Sky’s the Limit: Leveraging Drone Technology in Infrastructure Projects ; Global Infrastructure Hub (GI Hub): Sydney, Australia, 2024.

[52] Riebo Technology. Drone Bridge Inspection: AI—Enhanced Solutions for Safer Bridges. Riebo Technology: Chengdu, China. Available online: https://en.riebotech.com/solutions/bridge—inspection/ (accessed on 15 December 2025).

[53] FEDS Drone Powered Solutions. Drone Integration in Smart Cities: Building Resilient Urban Ecosystems. Available online: https://www.feds.group/blog/drone—integration—in—smart—cities—building—resilient—urban—ecosystems (accessed on 17 December 2025).

[54] Ali QI. Realization of a robust fog—based green VANET infrastructure. IEEE Syst. J. 2022, 17, 2465-2476. DOI: 10.1109/JSYST.2022.3215845

[55] Vigoroso M. AI Era Ushers in Drones as Infrastructure Inspectors and Algorithms as Failure Predictors. ERP Today, 4 June 2025. Available online: https://erp.today/ai—era—ushers—in—drones—as—infrastructure—inspectors—and—algorithms—as—failure—predictors/ (accessed on 17 December 2025).

[56] Alsharbaty FS, Ali QI. Smart electrical substation cybersecurity model based on WPA3 and cooperative hybrid intrusion detection system (CHIDS). Smart Grids Sustain. Energy 2024, 9, 11. DOI: 10.1007/s40866—024—00192—7

[57] Muñoz de Rivera González M. AI Workforce: Using AI and Drones to Simplify Infrastructure Inspections. AWS Machine Learning Blog, Amazon Web Services, 3 April 2025. Available online: https://aws.amazon.com/blogs/machine—learning/ai—workforce—using—ai—and—drones—to—simplify—infrastructure—inspections/ (accessed on 19 December 2025).

[58] PwC. Clarity from Above: PwC Global Report on the Commercial Applications of Drone Technology ; PwC Drone Powered Solutions: Warsaw, Poland, 2016.

[59] Equinox’s Drones. Drones in Smart City Planning/Urban Planning: A Bird's—Eye View. Available online: https://equinoxsdrones.com/drones—in—smart—city—planning—urban—planning—a—birds—eye—view/ (accessed on 20 December 2025).

[60] Flyby Guys. Smart City Traffic Management with Drone Technology. Available online: https://www.flybyguys.com/blog/monitoring—traffic—drones (accessed on 21 December 2025).

[61] Ullah I, Singh SK, Adhikari D, Khan H, Jiang W, Bai X. Multi—Agent Reinforcement Learning for task allocation in the Internet of Vehicles: Exploring benefits and paving the future. Swarm Evol. Comput. 2025, 94, 101878. DOI: 10.1016/j.swevo.2025.101878

[62] Lui S. Eye in the Sky: Transforming Mobility with Drones and AI. Digital AECOM, AECOM, 8 May 2025. Available online: https://digital.aecom.com/article/eye—in—the—sky/ (accessed on 21 December 2025).

[63] Greenroads.ai. Drone Analytics for Smart Cities. Available online: https://www.greenroads.ai/drone—analytics/ (accessed on 21 December 2025).

[64] Manzini T, Murphy R. CLARKE: Computer Vision and Learning for Analysis of Roads and Key Edifices. Texas A&M University Stories, Texas A&M University, 28 July 2025. Available online: https://stories.tamu.edu/news/2025/07/28/ai—turns—drone—footage—into—disaster—response—maps—in—minutes/ (accessed on 16 December 2025).

[65] Folio3 AI. Real—Time Disaster Response with AI Drone Swarms. Available online: https://www.folio3.ai/blog/ai—drone—swarms—disaster—response/ (accessed on 22 December 2025).

[66] Qassab MS, Ali QI. A UAV—based portable health clinic system for coronavirus hotspot areas. Healthc. Technol. Lett. 2022, 9, 77-90. DOI: 10.1049/htl2.12035

[67] SafeSight Exploration. The Role of AI and Autonomous Drones in Disaster Management. Available online: https://safesightxp.com/2024/12/23/the—role—of—ai—and—autonomous—drones—in—disaster—management/ (accessed on 22 December 2025).

[68] Butt AUR, Saba T, Khan I, Mahmood T, Khan AR, Singh SK, et al. Proactive and data—centric internet of things—based fog computing architecture for effective policing in smart cities. Comput. Electr. Eng. 2025, 123, 110030. DOI: 10.1016/j.compeleceng.2024.110030

[69] Hello Future (Orange). Search and Rescue: Drones That Detect Human Voices Under Collapsed Buildings. Hello Future, Orange, 6 June 2024. Available online: https://hellofuture.orange.com/en/search—and—rescue—drones—that—detect—human—voices—under—collapsed—buildings/ (accessed on 22 December 2025).

[70] McNabb M. Drones in Disaster Response: Harnessing the Potential of Unmanned Aerial Systems. DRONELIFE, 16 January 2024. Available online: https://dronelife.com/2024/01/16/drones—in—disaster—response—harnessing—the—potential—of—unmanned—aerial—systems/ (accessed on 22 December 2025).

[71] Shi L, Yan S, Li W. Consensus and products of substochastic matrices: Convergence rate with communication delays. IEEE Trans. Syst. Man Cybern. Syst. 2025, 55, 4752-4761. DOI: 10.1109/TSMC.2025.3559667

[72] Li W, Yan S, Shi L, Yue J, Shi M, Lin B, et al. Multiagent consensus tracking control over asynchronous cooperation—Competition networks. IEEE Trans. Cybern. 2025, 55, 4347-4360. DOI: 10.1109/TCYB.2025.3583387

[73] Vaimanika Aerospace. Smart Cities Powered by Drones: Redefining Urban Development. Available online: https://vaimanikaaerospace.com/smart—cities—powered—by—drones—redefining—urban—development/ (accessed on 23 December 2025).

[74] Gjondeva A. The Role of Drones in Enhancing Smart City Development. Supply Chain Game Changer™, 24 July 2025. Available online: https://supplychaingamechanger.com/the—role—of—drones—in—enhancing—smart—city—development/ (accessed on 23 December 2025).

[75] Ali QI. Design & implementation of a mobile phone charging system based on solar energy harvesting. In Proceedings of the 2010 1st International Conference on Energy, Power and Control (EPC—IQ), Basrah, Iraq, 30 November—2 December 2010; pp. 264-267.

[76] Ultralytics. YOLOv8 on NVIDIA Jetson Using TensorRT. Available online: https://docs.ultralytics.com/guides/nvidia—jetson/ (accessed on 15 December 2025).

[77] Coral G. Edge TPU Performance Benchmarks. Available online: https://coral.ai/docs/edgetpu/benchmarks/ (accessed on 16 December 2025).

[78] Federal Aviation Administration (FAA). Operation of Small Unmanned Aircraft Systems Over People. Final Rule. Federal Register 2021, 86, 4314-4445. Docket No. FAA—2018—1087. Available online: https://www.federalregister.gov/documents/2021/01/15/2020—28947/operation—of—small—unmanned—aircraft—systems—over—people (accessed on 15 December 2025).

[79] European Union Aviation Safety Agency (EASA). Easy Access Rules for Unmanned Aircraft Systems (Regulations (EU) 2019/947 and 2019/945). Available online: https://www.easa.europa.eu/en/document—library/easy—access—rules/easy—access—rules—unmanned—aircraft—systems—regulations—eu (accessed on 18 December 2025).

[80] Stanczak J. NR Support for UAVs. 3GPP Highlights, Issue 06, 3rd Generation Partnership Project (3GPP), July 2023. Available online: https://www.3gpp.org/technologies/nr—uav (accessed on 18 December 2025).

[81] Ali QI. Design and implementation of an embedded intrusion detection system for wireless applications. IET Inf. Secur. 2012, 6, 171-182. DOI: 10.1049/iet—ifs.2010.0245

[82] Ali QI. Securing solar energy—harvesting road—side unit using an embedded cooperative—hybrid intrusion detection system. IET Inf. Secur. 2016, 10, 386-402. DOI: 10.1049/iet—ifs.2014.0456

[83] Haider ZA, Ullah I, Abdusalomov A, Shah M, Khan MZ, Zneid BA. Edge—intelligent semantic aggregation in blockchain—secured 6G UAV—assisted Internet of vehicles. J. Electron. Sci. Technol. 2026, 24, 100350. DOI: 10.1016/j.jnlest.2026.100350

[84] Enterprise D. Matrice 350 RTK Specifications. Available online: https://enterprise.dji.com/matrice—350—rtk/specs (accessed on 14 December 2025).

[85] Skydio. Skydio X10 Technical Specifications. Available online: https://www.skydio.com/x10/technical—specs (accessed on 15 December 2025).

[86] Robotics A. EVO Max 4T V2 Specifications. Available online: https://www.autelrobotics.com/products/evo—max—4t/ (accessed on 16 December 2025).

[87] Azure M. Azure Digital Twins Documentation. Available online: https://docs.microsoft.com/azure/digital—twins/ (accessed on 13 December 2025).

[88] Ullah I, Noor A, Abbas M, Garg S, Choi BJ, Hassan MM, et al. Optimizing smart city services by utilizing appropriate characteristics of digital twin for urban excellence. Alex. Eng. J. 2025, 122, 399-410. DOI: 10.1016/j.aej.2025.02.085

[89] ETSI. Context Information Management (CIM); NGSI—LD API. ETSI GS CIM 009 V1.8.1. Available online: https://www.etsi.org/deliver/etsi_gs/CIM/001_099/009/01.08.01_60/gs_cim009v010801p.pdf (accessed on 15 December 2025).

[90] Liang S, Khalafbeigi T, van Der Schaaf H, Miles B, Schleidt K, Grellet S, et al. Ogc sensorthings api part 1: Sensing version 1.1. 2021. Available online: https://brgm.hal.science/hal—04235851/ (accessed on 18 December 2025).

[91] ASTM F3269—21; Standard Practice for Methods to Safely Bound Behavior of Aircraft Systems Containing Complex Functions Using Run—Time Assurance . ASTM International: West Conshohocken, PA, USA, 2021.

[92] ASTM F3411—22a; Standard Specification for Remote ID and Tracking . ASTM International: West Conshohocken, PA, USA, 2022.

[93] European Commission. Commission implementing regulation (EU) 2021/664 of 22 April 2021 on a regulatory framework for the U—Space (text with EEA relevance). Off. J. Eur. Union 2021, 139, 161-183. Available online: https://eur—lex.europa.eu/legal—content/EN/TXT/PDF/?uri=CELEX:32021R0664 (accessed on 18 December 2025).

[94] Raman R, Datta U. The role of ‘Unmanned Aerial Vehicles’ in smart city planning and management. In Proceedings of the International Conference on Unmanned Aerial System in Geomatics, Roorkee, India, 2—4 April 2021; pp. 99-120.

[95] Bai X, Jiang H, Li C, Ullah I, Al Dabel MM, Bashir AK, et al. Efficient hybrid multi—population genetic algorithm for multi—UAV task assignment in consumer electronics applications. IEEE Trans. Consum. Electron. 2025, 71, 2395-2406. DOI: 10.1109/TCE.2025.3563339

[96] Arshi O, Khan IU, Kaushik K, Iqbal N, Ullah I, Slimani K. Unmanned aerial vehicles swarm for protecting smart cities. In Smart Sustainable Cities: Concepts and Applications Publ ; Springer: Berlin/Heidelberg, Germany, 2024; pp. 222-239. DOI: 10.1007/979—8—8688—1047—3