sMaRT

Inspiration

COVID-19 has a huge impact on public transportation. The MRT operators announced in April that they would adjust operation plans, and they did reduce the number of trains in operation. However, we believe that the adjustment on operations should be more sophisticated and adaptive considering the evolving pandemic situation.

What it does

Our model takes the severity of areas with MRT stations into consideration. These severity indexes can provide an insight into the number of passengers getting on and off the train. We then have an objective function that seeks to balance and maximize the efficiency of traveling as well as the cost-effectiveness of train operations.

How I built it

We built an objective function with aspects including but not limited to travel efficiency, rate of occupancy, and average waiting time, and we assigned coefficients according to their importance (in our example, we seek to balance the objectives). Assuming that we have the relevant datasets, we used linear regression, decision tree, and reinforcement learning to find the final output.

Challenges I ran into

The reinforcement learning part was very tricky in coding. We tried our best but did not finish the code. Problem identification and modeling were difficult. It took us three hours to figure out a sophisticated and comprehensive model.

Accomplishments that I'm proud of

We finally figured out the objective function and it was an ingenious design in our opinion.

What I learned

It is always good to start with what we can do and have a clear overview of the project.

What's next for sMaRT

Acquire datasets and develop relevant coding parts. After that, we should put it to testing.

Built With

• jupyter
• python