Results

Deliverable

DeliverableWorkpackageResponsableTypeAccess typeDue date
D5.1 Consortium AgreementWP5Răzvan Gabriel BobocDocumentPrivateM2
10.2025
D4.1 Communication PlanWP4Elena NaeReportPublicM4
12.2025
D4.2 Dissemination and Exploitation Plan (M4, M24)WP4Elena NaeReportPublicM4
12.2025
D5.3 Yearly scientific/technical and financial report to UEFISCDIWp5Răzvan Gabriel Boboc
Eugen Valentin Butilă
ReportPublicM4
12.2025
D2.1 Public survey and stakeholder engagement reportWP2Daniel VoineaReportPublicM14
10.2026
D1.1 Report on AV simulation model with DTAWP1Dinu CovaciuReportPublicM16
12.2025
D5.3 Yearly scientific/technical and financial report to UEFISCDIWP5Răzvan Gabriel Boboc
Eugen Valentin Butilă
ReportPublicM16
12.2025
D1.2 Report on scenario analysisWP1Dinu CovaciuReportPublicM24
08.2027
D2.2 Strategic roadmap for the gradual integration of AVsWP2Daniel VoineaReportPublicM24
08.2027
D3.1 Knowledge transfer and best practices reportWP3Eugen Valentin Butilă & Grigore AmbrosiReportPublicM24
08.2027
D4.2 Dissemination and Exploitation PlanWP4Elena NaeReportPublicM24
08.2027
D4.3 Project proposal submission proofWP4Elena NaeReportPublicM24
08.2027
D5.3 Yearly scientific/technical and financial report to UEFISCDIWP5Răzvan Gabriel Boboc
Eugen Valentin Butilă
ReportPublic
M24
08.2027

Technological Innovations

1. Edge-Based AI Traffic Monitoring System

The AVSIM system is designed as an end-to-end pipeline deployed at the network edge.

Hardware platform

  • NVIDIA Jetson AGX Orin (embedded GPU computing)
  • Stereo camera ZED 2 (RGB + depth)
  • On-device processing with TensorRT-optimized models

AI models

  • Object detection: YOLOv8 / EfficientDet-Lite
  • Multi-object tracking: DeepSORT / ByteTrack
  • 3D localization: Depth estimation + camera calibration
  • Privacy protection: Real-time face/body blurring or pixelation

Key capabilities

  • Real-time detection of vehicles & pedestrians (10–20 FPS on Orin)
  • 3D trajectory reconstruction
  • Volume counting based on virtual lines/zones
  • Extraction of speed, headway, densities, and turning movements
  • Anonymized, metadata-only data export compliant with GDPR principles

2. Traffic Simulation – Dynamic Traffic Assignment (DTA)

Simulation work focuses on understanding how AV penetration affects:

  • congestion,
  • queue formation and dissipation,
  • route choice behavior,
  • multimodal interactions.

Software tools

  • PTV VISSIM (microscopic simulation)
  • PTV VISUM / external solvers (macroscopic/mesoscopic DTA)
  • Hybrid Visum DTA + Vissim microsimulation workflow

DTA features implemented

  • Time-dependent demand
  • Multi-class modelling (Autonomous Vehicles vs Human-driven Vehicles)
  • AV behavioural parameters: shorter headways, smoother acceleration, cooperative lane changing
  • Convergence evaluation and network loading stability tests

Public Acceptance Study

A structured survey assesses:

  • perceived usefulness
  • ease of use
  • technological anxiety
  • perceived risks
  • trust in AV systems
  • social influence
  • behavioural intention to use Avs

Survey pathways differ for respondents with and without prior experience using ADAS or autonomous functions.

Progress in the First 4 Months

WP1 – Simulation & Data Collection (Preparatory Work)

  • Selection of pilot intersections (Brașov: Civic Centre Ring, Chișinău: Petricani–Balcani Intersection)
  • Calibration of edge devices (Jetson Orin + ZED 2)
  • Definition of the AI data pipeline
  • Collaboration protocol with Brașov City Hall & Traffic Monitoring Centre

WP2 – Public Acceptance

  • Full questionnaire structure
  • Mapping of stakeholder groups in both cities
  • Identification of key psychological constructs

WP3 – Knowledge Transfer

  • First research mobility of UTM team to UTBV (4 researchers)
  • Technical workshop on PTV VISSIM
  • Creation of first demonstration simulation models

WP4 – Communication & Dissemination

  • Deliverables completed:
    • D4.1 – Communication Plan
    • D4.2 – Dissemination & Exploitation Plan
  • Media visibility in local press, university communication channels, and social media
  • Presentation at AMMA 2025 (Cluj-Napoca)

WP5 – Management

  • Consortium Agreement signed
  • Project kick-off meeting in Brașov
  • Procedures for reporting, scheduling, and data management established
  • Visits to the Traffic Monitoring Centre and Research Institute

Key Achievements (Months 1–4)

  • 3 technical deliverables submitted
  • 1 research mobility completed
  • 1 workshop organised
  • 1 international conference presentation
  • Project website under development (80%)
  • Two public announcements & media coverage