Imagine a city where traffic flows smoothly, and passengers wait for a shorter time. Sounds like a dream, right? A Reddit user is trying to make this a reality by simulating a city, traffic, and routes using SUMO software. The goal is to reduce waiting time for passengers and maximize profits for the organization. But, is this project doable?
The project involves five steps: simulating the city, getting data from SUMO using Python, feeding the data to a Graph Neural Network (GNN), training the GNN to make predictions using a Reinforcement Learning (RL) agent, and using the decisions of the RL agent in SUMO. The objectives are clear, but there are potential errors and challenges. For instance, the model may not work well in the real world due to factors like accidents or riots. The passengers’ predicting model could also go wrong, and the RL agent might make reward-giving decisions that aren’t preferred.
The biggest challenge, however, is that the team has no experience with SUMO, Python, GNN, and RL. They’re preparing for a tough journey ahead, but it’s worth exploring the possibilities of AI in improving our daily lives.
What do you think? Can AI really make a difference in reducing waiting time and boosting profits? Share your thoughts!