Kaitlyn’s Status Report for 3/16/24

Work Done

This week I spent a lot of time working on the Ethics Assignment and also finishing up the simulation code to use in our ML model. At the moment, I am able to remotely run a SUMO simulation in Python using TraCI. I currently have functions that can retrieve the waiting times for cars however I realized I need to modify my simulation to add lane area detectors to be able to tell which cars are in what lane and calculate the wait time at the light correctly since there is no other way to locate the cars in relation to a light object without additional objects in the simulation. This should not be too difficult now that I found out this feature exists in SUMO, however it will take some time to configure, since it seems that I have to add it in via code rather than using the SUMO GUI.

This is what the current simulation looks like:

I was able to simplify the simulation by removing buildings, so that the simulation is easier to manage and is more efficient to run. The simulation pictured is all controlled by TraCI and each step is triggered by the code.

Schedule

I should be able to finish the SUMO and ML by the time we do our interim demo, since the bulk of the time is the simulation, which I have been making good progress towards and should be done with by tomorrow. Unfortunately, this is a little behind schedule since SUMO has been quite difficult to work on due to the lack of documentation and use by developers. I have had to get help from random Stack Overflow questions, however there are very few trying to do similar work as us, so it takes longer to debug and implement.

I think once this is over, the machine learning part should not be too long to finish since there is more coverage on PyTorch and the ML models than SUMO and TraCI.

Tasks This Week

• Finish up SUMO simulation (actually this time)
• Make a basic ML model implementing simulation – will optimize hyperparameters in upcoming weeks