A python package that computes an optimal motion plan for approaching a red light

Overview

Code style: black License: AGPL v3

redlight_approach

redlight_approach is a Python package that computes an optimal motion plan during traffic light approach.

RLA_demo.mov

Given the parameters of the road and vehicle, and a probability distribution describing when the traffic light will turn green, redlight_approach finds the optimal motion plan to minimize the expected amount of time spent traversing the intersection. It enacts the motion plan in a SUMO simulation, and a standard human driver is simulated for comparison. It reports the difference in time between the vehicles. The simulation above is cherry-picked, but typical use involves running many simulations with the red light duration sampled from the probability distribution. Preliminary findings show that this traffic light approach planner will save vehicles time in realistic scenarios.

Next Steps: Baysian Updating from World State

Currently, redlight_approach uses a fixed green light event probability distribution throughout a single approach. It's clear this is naive, as there is information about the traffic light cycle available through observation of the world state. For instance, seeing a vehicle slow down that is approaching the intersection perpendicular to your approach, indicates that their traffic light is no longer green. This suggests that your light may turn green soon, depending on the traffic light cycle. There are other possible information sources, like direct observation of the light from another vehicle of a connected fleet. Updating the distribution from an observation of the world during approach would increase the performance of the system.

Formally, the goal of this part of the project is to map world state to a Baysian update of the probability distribution. Specifically, this will involve using a neural net to map an aspect of world state, like the position of other vehicles, to a likelyhood function defined over the support of the green light event. This requires a dataset which will be generated by running many simulations. This effort is currently underway in the baysian_update branch.

Requirements

  1. Linux or macOS

  2. Miniconda or Anaconda
    To install, visit Conda Installation

  3. SUMO: Installation instructions for Linux (Ubuntu) and macOS

Install SUMO (Ubuntu)

  1. Build SUMO from source (see SUMO Linux Build for more details)
sudo apt-get install git cmake python3 g++ libxerces-c-dev libfox-1.6-dev libgdal-dev libproj-dev libgl2ps-dev
git clone --recursive https://github.com/eclipse/sumo
export SUMO_HOME="$PWD/sumo"
mkdir sumo/build/cmake-build && cd sumo/build/cmake-build
cmake ../..
make -j$(nproc)
sudo make install

Install SUMO (macOS)

  1. Install Homebrew if you don't have it:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
  1. Install XQuartz with Homebrew:
brew install --cask xquartz
  1. Install SUMO with Homebrew:
brew tap dlr-ts/sumo
brew install sumo

Installation of redlight_approach

  1. Edit .bashrc or .zshrc:
  • Add these lines to your shell's config file:

    # Your .bashrc or .zshrc file
    
    export SUMO_HOME="/path/to/sumo"
    export PYTHONPATH="$PYTHONPATH:/path/to/parent/"
    
  • Replace /path/to/sumo above with your sumo location, which you can find with which sumo.

  • Replace /path/to/parent above with the directory into which you clone this repo, which you can find with pwd.

  • Load these environment variables with

    source ~/.bashrc
    # or 
    source ~/.zshrc
    
  1. Clone this repository:
git clone https://github.com/basilforlife/redlight_approach.git
  1. Change directories to the root of redlight_approach:
cd redlight_approach
  1. Create and activate conda env:
conda env create -f environment.yml
conda activate rla
  1. If you're going to contribute, add pre-commit hooks:
pre-commit install

Usage

Confirm installation was successful with the test suite:

pytest

Typical Use

Run the default scenario with the -g (graphical) option:
Note: on macOS, XQuartz must be running in order to use the graphical option. Start XQuartz from the application folder.

python simple_comparison.py -c parameter_files/original.json -g

To speed up future runs use the the -p option:

python simple_comparison.py -c parameter_files/original.json -p original.pickle

On subsequent runs, use -u to load the same configuration as before:

python simple_comparison.py -u original.pickle

To plot the result of N runs, use the -N option:

python simple_comparison.py -u original.pickle -N 100

For a complete list of options, use the -h option:

python simple_comparison.py -h

License

This work is licensed under the GNU Affero General Public License v3.0. Feel free to contact me if you have any questions about the project.

Owner
Jonathan Roy
Jonathan Roy
Small projects for python beginners.

Python Mini Projects For Beginners I recently started doing the #100DaysOfCode Challenge in Python. I've used Python before, but I had switched to JS

Sreekesh Iyer 10 Dec 12, 2022
Esercizi di Python svolti per il biennio di Tecnologie Informatiche.

Esercizi di Python Un piccolo aiuto per Sofia che nel 2° quadrimestre inizierà Python :) Questo repository (termine tecnico di Git) puoi trovare tutti

Leonardo Essam Dei Rossi 2 Nov 07, 2022
🍕 A small app with capabilities ordering food and listing them with pub/sub pattern

food-ordering A small app with capabilities ordering food and listing them. Prerequisites Docker Run Tests docker-compose run --rm web ./manage.py tes

Muhammet Mücahit 1 Jan 14, 2022
This is the community maintained fork of ungleich's cdist (after f061fb1).

cdist This is the community maintained fork of ungleich's cdist (after f061fb1). Work is split between three repositories: cdist - implementation of t

cdist community edition 0 Aug 02, 2022
This is the improvised version of Dobot Magician which can be implemented for Dobot M1

pydobotM1 This is the edited driver for Dobot M1 version of the original pydobot library intended for use with the Dobot Magician. Here's what you nee

Shaik Abdullah 2 Jul 11, 2022
A collection of resources on neural rendering.

awesome neural rendering A collection of resources on neural rendering. Contributing If you think I have missed out on something (or) have any suggest

1.8k Dec 30, 2022
Simple programming language built on Python.

Serial Another programming language. Built on Python. Building and running program In order to run the program on serial, unfortunately you still need

Aleksey Demchenkov 1 Dec 09, 2021
Python-Kite: Simple python code to make kite pattern

Python-Kite Simple python code to make kite pattern. Getting Started These instr

Anoint 0 Mar 22, 2022
This is the repo for Uncertainty Quantification 360 Toolkit.

UQ360 The Uncertainty Quantification 360 (UQ360) toolkit is an open-source Python package that provides a diverse set of algorithms to quantify uncert

International Business Machines 207 Dec 30, 2022
tox-gh is a tox plugin which helps running tox on GitHub Actions with multiple different Python versions on multiple workers in parallel

tox-gh is a tox plugin which helps running tox on GitHub Actions with multiple different Python versions on multiple workers in parallel. This project is inspired by tox-travis.

tox development team 19 Dec 26, 2022
A password genarator/manager for passwords uesing a pseudorandom number genarator

pseudorandom-password-genarator a password genarator/manager for passwords uesing a pseudorandom number genarator when you give the program a word eg

1 Nov 18, 2021
Py4J enables Python programs to dynamically access arbitrary Java objects

Py4J Py4J enables Python programs running in a Python interpreter to dynamically access Java objects in a Java Virtual Machine. Methods are called as

Barthelemy Dagenais 1k Jan 02, 2023
This is a practice on Airflow, which is building virtual env, installing Airflow and constructing data pipeline (DAGs)

airflow-test This is a practice on Airflow, which is Builing virtualbox env and setting Airflow on that env Installing Airflow using python virtual en

Jaeyoung 1 Nov 01, 2021
This is a program for Carbon Emission calculator.

Summary This is a program for Carbon Emission calculator. Usage This will calculate the carbon emission by each person on various factors. Contributor

Ankit Rane 2 Feb 18, 2022
Automator anble you to create automations on your system

WELCOME TO AUTOMATOR BETA This programm is able to create automations on your system. This programm is only an experimantal release; infact it works v

Davide 1 Jan 12, 2022
Demo content - Automate your automation!

Automate-AAP2 Demo Content - Automate your automation! A fully automated Ansible Automation Platform. Context Installing and configuring Ansible Autom

0 Oct 27, 2022
Annotates sequences with Eggnog-mapper and hhblits against PDB70

Annotating "hypothetical" proteins with the PDB See config/ for configuration information. This workflow takes as input a set of protein sequences. It

1 Apr 05, 2022
A corona information module

A corona information module

Fayas Noushad 3 Nov 28, 2021
A Klipper plugin for accurate Z homing

Stable Z Homing for Klipper A Klipper plugin for accurate Z homing This plugin provides a new G-code command, STABLE_Z_HOME, which homes Z repeatedly

Matthew Lloyd 24 Dec 28, 2022
run-js Goal: The Easiest Way to Run JavaScript in Python

run-js Goal: The Easiest Way to Run JavaScript in Python features Stateless Async JS Functions No Intermediary Files Functional Programming CommonJS a

Daniel J. Dufour 9 Aug 16, 2022