VRU-aware GLOSA

This repository contains the simulation implementation for the paper: "VRU-Aware GLOSA: Integrating VRUs into Green Light Optimized Speed Advisory for Right-Turn Conflicts at Signalized Intersections", to be presented at SUMO User Conference 2026

This work evaluates the effect of GLOSA and VRU-aware GLOSA on traffic efficiency and safety at a signalised junction in Ingolstadt, Germany, using the SUMO traffic simulator. Three scenarios are compared: a baseline (no control), standard GLOSA, and a VRU-aware GLOSA extension that detects approaching cyclists and adjusts the speed advisory for right-turning vehicles to prevent conflicts.

---

Installation

Prerequisites

• Anaconda or Miniconda
• SUMO traffic simulator version 1.23.0 (results may vary with other versions)

Cloning the repository

git clone https://github.com/mahajanpushkar/VRU_Aware_GLOSA.git
cd VRU_Aware_GLOSA

Setting up the environment

Create and activate a Conda environment using the provided environment.yml:

conda env create -f environment.yml
conda activate vru_aware_glosa

SUMO Python path

Ensure the SUMO tools directory is on your PYTHONPATH so that traci can be imported:

# Linux / macOS
export PYTHONPATH="$SUMO_HOME/tools:$PYTHONPATH"

# Windows (PowerShell)
$env:PYTHONPATH = "$env:SUMO_HOME\tools;$env:PYTHONPATH"

---

Input Files

The SUMO network and route files are included in the repository under sumo_files/.

sumo_files/
├── ingolstadt_tiny.sumocfg
├── net/
│   └── ingolstadt_tiny.net.xml
└── routes/
    ├── ingolstadt_tiny_vehicles_12H.rou.xml
    ├── ingolstadt_tiny_bikes_12H.rou.xml
    ├── ingolstadt_tiny_vehicles_1H.rou.xml
    ├── ingolstadt_tiny_bikes_1H_Conv_above_60.rou.xml
    ├── ingolstadt_tiny_bikes_1H_Conv_below_40.rou.xml
    ├── ingolstadt_tiny_bikes_1H_Conv_between_40_and_60.rou.xml
    ├── ingolstadt_tiny_bikes_1H_Ebike_above_60.rou.xml
    ├── ingolstadt_tiny_bikes_1H_Ebike_below_40.rou.xml
    ├── ingolstadt_tiny_bikes_1H_Ebike_between_40_and_60.rou.xml
    ├── ingolstadt_tiny_bikes_1H_SPedelec_above_60.rou.xml
    ├── ingolstadt_tiny_bikes_1H_SPedelec_below_40.rou.xml
    └── ingolstadt_tiny_bikes_1H_SPedelec_between_40_and_60.rou.xml

---

Configuration

All parameters are centralised in config.py. Before each run, set the four variables at the top of the file to select the study, bicycle demand variant, and simulation time window:

ACTIVE_STUDY   = "A"             # "A" = 12-hour study  |  "B" = 1-hour study (11:00–12:00)
ACTIVE_VARIANT = "Conv_above_60" # Study B only — ignored when ACTIVE_STUDY = "A"
SIM_BEGIN      = 0               # Study A: 0  |  Study B: 39600
SIM_END        = 43200           # Study A: 43200  |  Study B: 43200

Study B supports nine bicycle variants across three bike types (Conventional, E-bike, S-Pedelec) and three rider age groups (below 40, between 40–60, above 60). No other file needs to be edited between runs — the correct network, route files, and output paths are resolved automatically.

---

Running the Simulations

Run all commands from the project root directory (the folder containing config.py).

# Baseline — no speed control
python -m simulations.baseline

# GLOSA — speed advisory for all vehicles
python -m simulations.glosa

# VRU-aware GLOSA — GLOSA with bicycle conflict detection
python -m simulations.vru_aware_glosa

To use the SUMO GUI, replace "sumo" with "sumo-gui" in the SUMO_CMD list at the top of the relevant simulation script.

---

Output Files

Each run writes three files to the corresponding scenario subfolder under outputs/:

outputs/
├── study_A/
│   ├── baseline/
│   │   ├── right_turn_metrics.csv
│   │   ├── safety_conflicts_ttc5_pet3.csv
│   │   └── sumo_log_baseline.txt
│   ├── glosa/
│   └── vru_aware_glosa/
└── study_B/
    └── <variant>/
        ├── baseline/
        ├── glosa/
        └── vru_aware_glosa/

Traffic-efficiency CSV — right_turn_metrics.csv

One row per right-turning motor vehicle that completed its trip.

Column	Description
veh_id	Vehicle identifier
depart_time_s	Departure time (s)
arrive_time_s	Arrival time (s)
travel_time_s	Total travel time (s)
waiting_time_s	Total time spent below the stop-speed threshold (s)
stops_count	Number of moving → stopped transitions
co2_mg	Total CO₂ emissions (mg)
co_mg	Total CO emissions (mg)
cox_mg	Combined COx = CO₂ + CO (mg)
fuel_mg	Total fuel consumption (mg)
electricity_Wh	Total electricity consumption (Wh)
red_waiting_time_s	Waiting time at red signals specifically (s)
red_stops_count	Number of stops at red signals
first_red_stop_start_s	Simulation time when the first red stop began (s)
first_red_restart_s	Simulation time of restart after the first red stop (s)
restart_edge_id	Road edge where the vehicle restarted after its first red stop

Safety CSV — safety_conflicts_ttc5_pet3.csv

One row per road user (vehicle or bicycle) that met the conflict filter: TTC ≤ 5 s and PET ≤ 3 s.

Column	Description
user_id	Road-user identifier
type	vehicle_right or bicycle_straight
depart_time_s	Departure time (s)
arrive_time_s	Arrival time (s)
travel_time_s	Total travel time (s)
zone_entered_t	Time of first entry into the conflict zone (s)
zone_exited_t	Time of first exit from the conflict zone (s)
min_ttc_s	Minimum Time-To-Collision recorded over the trip (s)
min_ttc_with	ID of the opposing road user at minimum TTC
min_pet_s	Minimum Post-Encroachment Time recorded over the trip (s)
min_pet_with	ID of the opposing road user at minimum PET

---

Project Structure

glosa_simulation/
├── config.py                        ← Study/variant selector + all shared settings
│
├── sumo_files/                      ← SUMO input files 
│   ├── ingolstadt_tiny.sumocfg
│   ├── net/
│   └── routes/
│
├── outputs/                         ← Auto-created at runtime; one folder per scenario
│
├── evaluators/
│   ├── traffic_efficiency.py        ← RightTurnEfficiencyEvaluator
│   └── safety_metrics.py            ← JunctionSafetyEvaluator (TTC & PET)
│
├── utils/
│   ├── network.py                   ← Lane lengths, distance-to-stop-line, route classification
│   ├── tls.py                       ← TLS cache and future signal state builder
│   └── glosa_core.py                ← GLOSA speed-advisory algorithm
│
├── simulations/
│   ├── baseline.py                  ← No speed control
│   ├── glosa.py                     ← GLOSA speed advisory
│   └── vru_aware_glosa.py           ← GLOSA + VRU bicycle conflict detection
│
├── .gitignore
└── README.md