Maps are very good at optimizing routes for cars. They use traffic data, congestion, road types, and real time conditions.

But with walking it's different.

A person usually walks at a constant speed. There are no traffic jams for walking. The only real thing that stops a person is traffic lights.

Because of the traffic lights, the shortest route may not always be the fastest one.

So the idea is simple: If a system knows your average walking speed, where traffic lights are, and how long red and green lights last, it can choose a path that reduces your waiting time.

This project assumes that all the required data is known, and implements the logic for time-based walking route optimization using traffic light cycles and constant human walking speed.

It uses Dijkstra’s algorithm with a wait time penalty at red lights to find the fastest path, and Breadth-first Search to simulate all paths for visualization.

We also assume that:

• person walks at a constant average speed
• traffic lights have fixed green and red cycles (no sensors or pedestrian buttons and etc.)
• route is flat (no stairs, slopes and etc.)
• person is a Law Abiding Citizen and follows traffic rules
• person does not start crossing the road if they cannot reach the other side before the light turns red (no crossing at the orange light)
• map graph is a Directed Acyclic Graph (for the sake of simplicity)

git clone https://github.com/Otina12/google-maps-optimized-walking
cd google-maps-optimized-walking

Windows PowerShell:

python -m venv venv
.\venv\Scripts\Activate.ps1

macOS or Linux:

python3 -m venv venv
source venv/bin/activate

pip install -r requirements.txt

From the project root:

cd src/demo
python main.py

Edit input/points.txt:

point,x,y
A,1420,1290
B,1440,1240

Edit input/edges.txt.

source,target,green_time,red_time,cycle_offset
A,B,20,60,0
A,E,0,0,0

Important: Edges should form a Directed Acyclic Graph.

Update images at these paths:

• map image
• walker icon

After running, the simulator generates:

• an mp4 video
• an animated webp

This project is a simplified model and does not fully represent real-world walking conditions.

Things not considered:

• traffic lights are often adaptive and not fixed cycle
• pedestrian phases are different from car green phases
• push button crossings and sensor based lights
• variable walking speed due to slope, stairs, surface, weather, or crowd density
• sidewalk quality, safety, lighting, and accessibility
• real world map inaccuracies and construction

Possible improvements:

• use probabilistic wait times instead of fixed waits
• add uncertainty and expected time instead of single time values
• add user preferences such as safety, accessibility, or fewer crossings

The example input uses a real map route in Tbilisi, Georgia from Point A to Point B. Traffic light positions, timings and cycles are close to actual values.

This is only a concept project. It shows that walking navigation can easily be improved.