A Comprehensive Survey on Infrastructure-Independent and Adaptive Ultrasonic Parking Slot Detection Systems

Abstract

Urban infrastructure frequently demonstrates vulnerability during extreme conditions and complicated parking scenarios, where depending on expensive LiDAR or vision-centric modules results in operational failures during severe weather or network blackouts. This study introduces an innovative, software-oriented framework formulated to close the divide between economical consumer hardware and highly accurate spatial recognition. The proposed model is driven by the Adaptive Spatial Positioning (ASP) algorithm, employing conventional ultrasonic transducers to secure exact boundary localization by evaluating multiple sound wave reflections. To counteract the \"ghost\" signals and acoustic interference common in ultrasonic devices, the architecture integrates an adaptive filtering protocol and a decision-tree fusion mechanism analyzing inputs from ipsilateral sensors. Engineered for maximum computational brevity, the ASP logic functions on a highly restricted memory footprint, rendering it exceptionally appropriate for standard microcontrollers. To corroborate the proposed architecture, a physical prototype encompassing four parking bays serves as a scaled proof-of-concept. It processes a variety of ultrasonic metrics to pinpoint physical obstacles alongside the intervening vacant areas. Final quantitative metrics concerning processing speed, memory utilization, and boundary detection precision will be explicitly measured following the conclusion of the hardware trials. By combining a localized ESP32 processor with Wi-Fi connectivity, a Backend API, a MySQL Database, and a live Web Dashboard, the system guarantees uninterrupted remote tracking and data archiving. This bypasses the need for resource-heavy, cloud-dependent control units for essential detection tasks. Ultimately, this inquiry provides a highly scalable, infrastructure-agnostic strategy for automated parking using widely available components.

Introduction

The text proposes “Smart Sonar Parking: Adaptive Ultrasonic Parking Slot Detection,” a low-cost, autonomous parking detection system designed to overcome the limitations of camera-, LiDAR-, and conventional sensor-based parking systems.

Problem

Modern parking systems often rely on computer vision or LiDAR, but these technologies can be:

• Expensive and computationally demanding.
• Dependent on GPUs or powerful hardware.
• Vulnerable to rain, fog, glare, darkness, and reflective or dark-colored vehicles.

Ultrasonic sensors are cheaper and more weather-resistant, but they normally produce overlapping and misleading echoes, making it difficult to accurately identify parking-slot boundaries.

Proposed Solution

The proposed system combines HC-SR04 ultrasonic sensors with an ESP32 microcontroller and introduces an Adaptive Spatial Positioning (ASP) algorithm.

The system:

1. Collects ultrasonic Time-of-Flight (ToF) measurements.
2. Converts the measurements into spatial coordinates.
3. Uses moving-average filtering to reduce sensor noise.
4. Applies a Reflection Guard to eliminate false reflections and outliers.
5. Uses ASP to reconstruct the surrounding parking geometry.
6. Identifies empty spaces and calculates their dimensions.
7. Uses decision-tree-based sensor fusion to improve detection reliability.
8. Sends validated parking information through Wi-Fi to a backend system for storage and visualization.

Literature Findings

Previous studies demonstrate high performance but have important limitations:

• IR-based systems: Affordable but vulnerable to sunlight and reflective surfaces.
• YOLO/image-processing systems: Highly accurate but computationally expensive and sensitive to weather.
• OCR-based systems: Enable automated parking allocation but require clear visual conditions.
• AI + IoT systems: Improve parking assistance but can require substantial bandwidth and processing.
• Ultrasonic fusion systems: Provide accurate geometry but may depend on expensive FPGA hardware.

The proposed approach aims to achieve a better balance between cost, reliability, processing requirements, and environmental robustness.

System Architecture

The prototype uses multiple HC-SR04 sensors controlled by an ESP32. The ESP32 performs the main detection and mapping locally, while IoT infrastructure is used mainly for remote monitoring and historical data storage.

The architecture follows:

HC-SR04 → ESP32 → ToF measurement → Filtering → ASP → Spatial mapping → Empty-space detection → Dimension calculation → Sensor fusion → Parking decision

Validated results are transmitted to a backend API and stored in a MySQL database, while a web dashboard provides remote visualization.

Main Advantages

The proposed system offers:

• Low-cost, commercially available hardware
• Operation in poor lighting and adverse weather
• Local/edge processing without dependence on external smart-city infrastructure
• Reduced false detections through adaptive filtering and sensor fusion
• Automatic calculation of parking-space dimensions
• Remote monitoring and historical database storage

Conclusion

Mitigating urban parking congestion and deploying dependable automated assistance are critical for the evolution of modern smart cities. Traditional spatial mapping and infrastructure-reliant sensor arrays are frequently cost-prohibitive, sensitive to environmental shifts, and vulnerable to systemic failure. While camera modules supply detailed visual data under perfect conditions and LiDAR maps highly accurate 3D topographies, both suffer from severe weaknesses tied to weather interference and massive processing requirements. To overcome these barriers, the proposed Smart Sonar Parking framework capitalizes on affordable ultrasonic sensors, the Adaptive Spatial Positioning (ASP) algorithm, decentralized local computing via an ESP32 chip, and a resilient Web/Database networking layer. By evaluating overlapping ultrasonic echoes onboard the vehicle, the model successfully isolates physical boundaries and available spaces without requiring supplementary visual inputs. Additionally, by recording detected bays, exact dimensions, viability statuses, and timestamps within a MySQL ecosystem, the architecture effectively links self-sufficient detection with comprehensive remote tracking. Finalized metrics regarding precision, processing speed, and network latency will be systematically extracted during the forthcoming experimental phase. Through persistent innovations in edge computing, vehicle-centric parking models are poised to become significantly more robust, accessible, and proactive.

References

Visualization,\" ICEI, Pune, 2026. [2] K. Nesa and M. A. Kader, \"Image Processing and IoT Based Smart Parking Slot Detection and Notification System,\" ECCE, Bangladesh, 2025. [3] N. Srivani, A. Prasanth, and K. Srija, \"An Efficient Automatic Car Parking Slot Allocation System,\" ICSC, Bengaluru, 2025. [4] H. Albatati, S. Meer, and R. Albehise, \"Quick Park: Integrated Sensor-Based Smart Parking System with Chatbot for Enhanced Urban Mobility and Traffic Management,\" ICAISC, Jeddah, 2025. [5] B. Reddy and V. Vasanth, \"SmartPark: Efficiency,\" ICSCNA, India, 2024. [6] M. C. Chinnaiah, G. Divya Vani, and D. Hari Krishna, \"Geometry-Based Parking Assistance Using Sensor Fusion for Robots With Hardware Schemes,\" IEEE Sensors Journal, vol. 24, no. 5, pp. 4550–4562, 2024. [7] Y. Liu, K. Shang, D. Qiao, and C. Zhou, \"An Adaptive Spatial Positioning Algorithm for Ultrasonic-Only Parking Slot Detection System,\" IEEE Transactions on Instrumentation and Measurement, vol. 75, 2026. [8] Y. Shao, P. Chen, and T. Cao,detection,\" in Proc. IEEE pp. 3378–3381, Jul. 2018.

Copyright

Copyright © 2026 Akshath V Nadig, Ajay Jayaram, Hruthik S, Pavan Kumar MK. 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.