Newcastle is located close to Sydney but doesn't take full advantage of that proximity. Living in Newcastle provides a more relaxed lifestyle and cheaper cost of living. As Newcastle develops as a city, it is imperative that residents can travel to and from Sydney CBD. The ability to travel between these two cities efficiently will drastically improve Newcastle's economic footprint.
What it does and how we built it
Our solution models all possible train routes between Newcastle and Sydney as a directed graph. This graph contains 8,841,761,993,739,701,954,543,616,000,000 possible routes. We have leveraged a popular graph traversal algorithm combined with a weighting scheme that seeks to minimize the time passengers spend commuting, and maximize the number of passengers and distance travelled per passenger. In order to develop this graph, we analysed a range of datasets provided by NSW Open Data.
Challenges we ran into
The number of possible routes was enormous. We do not currently have the computational ability to brute force this problem and as a result we had to leverage a more efficient algorithm. Datasets are incomplete. Opal touch on data which is unmatched to touch off data makes for a less accurate analysis. As datasets of a higher quality are made available our algorithm will be able to render more optimal solutions.
Accomplishments that we're proud of
The algorithm returns the optimal result almost instantly - average powered hardware efficiently.
What we learned
We learned that the complexity of train scheduling is incredibly high.
What's next for NEWnet
Expanding the model to integrate more extensive datasets and applying it to other train networks.