Edge-AI Multi-Sensor Fusion Architecture for Barrier-Less Multi-Lane Free-Flow (MLFF) Electronic Toll Collection: Resolving Occlusion, OCR Degradation, and Class-Mismatch Leakage on High-Speed Corridors

The global transition from physical boom-barrier plazas to barrier-less Multi-Lane Free-Flow (MLFF) Electronic Toll Collection (ETC) is vital for mitigating severe highway congestion on high-density corridors. Pilot deployments planned across high-volume transit routes including Choryasi on NH-48 in Gujarat, Daulatpura, Manoharpur, and Shahjahanpur in Rajasthan, Mundka on UER-II, and Gharaunda on NH-44 in NCR—demonstrate that legacy single-modality architectures fail to maintain operational integrity at scale. While standalone Ultra-High Frequency (UHF) RFID and optical Automatic Number Plate Recognition (ANPR) achieve nominal performance baselines of 96%–98% and 75%–88% respectively under controlled conditions, the residual 1%–1.5% transaction failure rate creates unrecoverable revenue leakage exceeding millions of dollars annually on expressways handling over 50,000 daily Passenger Car Units (PCUs). This paper proposes a deterministic, Edge-AI multi-sensor fusion architecture operating on an IEEE 1588 Precision Time Protocol (PTP) synchronized network topology. The proposed system tightly integrates 3D Time-of-Flight (ToF) LiDAR point clouds, global-shutter 4K optical sensors paired with synchronized 850 nm pulsed infrared illumination, and Angle-of-Arrival (AoA) phase-array UHF RFID localization. By executing a 3D-to-2D Bird's-Eye-View (BEV) and pinhole perspective transformation pipeline, the architecture extracts spatial Regions-of-Interest (ROIs) for Spatial Transformer Network (STN) plate rectification, achieving sub-50 ms local edge inference, resolving multi-lane physical occlusions, and eliminating class-mismatch toll fraud.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-18
DOI
https://doi.org/10.5281/zenodo.24339706
Primary Topic
Autonomous Vehicle Technology and Safety
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Edge-AI Multi-Sensor Fusion Architecture for Barrier-Less Multi-Lane Free-Flow (MLFF) Electronic Toll Collection: Resolving Occlusion, OCR Degradation, and Class-Mismatch Leakage on High-Speed Corridors

Devansh Katole
Zenodo (CERN European Organization for Nuclear Research)
Autonomous Vehicle Technology and Safety
preprint

Edge-AI Multi-Sensor Fusion Architecture for Barrier-Less Multi-Lane Free-Flow (MLFF) Electronic Toll Collection: Resolving Occlusion, OCR Degradation, and Class-Mismatch Leakage on High-Speed Corridors

Devansh Katole
preprint en

Abstract

The global transition from physical boom-barrier plazas to barrier-less Multi-Lane Free-Flow (MLFF) Electronic Toll Collection (ETC) is vital for mitigating severe highway congestion on high-density corridors. Pilot deployments planned across high-volume transit routes including Choryasi on NH-48 in Gujarat, Daulatpura, Manoharpur, and Shahjahanpur in Rajasthan, Mundka on UER-II, and Gharaunda on NH-44 in NCR—demonstrate that legacy single-modality architectures fail to maintain operational integrity at scale. While standalone Ultra-High Frequency (UHF) RFID and optical Automatic Number Plate Recognition (ANPR) achieve nominal performance baselines of 96%–98% and 75%–88% respectively under controlled conditions, the residual 1%–1.5% transaction failure rate creates unrecoverable revenue leakage exceeding millions of dollars annually on expressways handling over 50,000 daily Passenger Car Units (PCUs). This paper proposes a deterministic, Edge-AI multi-sensor fusion architecture operating on an IEEE 1588 Precision Time Protocol (PTP) synchronized network topology. The proposed system tightly integrates 3D Time-of-Flight (ToF) LiDAR point clouds, global-shutter 4K optical sensors paired with synchronized 850 nm pulsed infrared illumination, and Angle-of-Arrival (AoA) phase-array UHF RFID localization. By executing a 3D-to-2D Bird's-Eye-View (BEV) and pinhole perspective transformation pipeline, the architecture extracts spatial Regions-of-Interest (ROIs) for Spatial Transformer Network (STN) plate rectification, achieving sub-50 ms local edge inference, resolving multi-lane physical occlusions, and eliminating class-mismatch toll fraud.

Zenodo (CERN European Organization for Nuclear Research)
Industry, innovation and infrastructure
Autonomous Vehicle Technology and Safety
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon