This project is just a tool to see how "green waves" (see ref. https://en.wikipedia.org/wiki/Green_wave) extraction works in traffic lights signal timing planning.

It includes calculation of best offset via genetic algorithm also.

Do not consider it to be any usable for production. It it just toy for understanding basics.

It works (as for 30th Jan 2026) with only single direction of the given road. UPD (31th Jan 2026): now both directions are possible to try to optimize

./docker.sh

docker run -p 36000:36000 greenwave

• Web UI: http://localhost:36000
• Swagger API docs: http://localhost:36000/api/greenwave/docs/

Forward direction only

Both directions

Note: In second example only reverse direction got "green wave" through all junctions, therefore vehicles per hour for forward direction is zero. So intensity is calculated only for cases when there ARE "green waves" which pass through all junctions.

docker run -p 35000:35000 \
    -e SERVER_HOST=0.0.0.0 \
    -e SERVER_PORT=35000 \
    -e SERVER_MAIN_PATH=api \
    -e SERVER_STARTUP_MESSAGE=true \
    -e USE_CORS=true \
    greenwave

Available environment variables:

* SERVER_MAIN_PATH: Web UI uses hardcoded path /api/greenwave. Changing this variable requires rebuilding the UI.

• Examples of JSON data for requests for can be found in the following README.md

docker pull ghcr.io/lddl/greenwave:latest
docker run -p 36000:36000 ghcr.io/lddl/greenwave:latest

docker pull dimahkiin/greenwave:latest
docker run -p 36000:36000 dimahkiin/greenwave:latest

• UI for OpenAPI Spec Documentations has been done via rapidoc

• JSON for Swagger has been prepared with swaggo/swag.

• REST API part is written with usage of labstack/echo.

• For logging rs/zerolog is used.

• BurntSushi/toml for TOML support
• stretchr/testify for testing tools