Solving this is a massive task. With just 10 passengers requesting at around the same time, you’ve got over 2,000 possible pickup and drop off sequences.

The challenge with other on-demand algorithms on the market is that they don't optimize in real-time, meaning - they respond too slowly to what's actually happening on the streets.

TRC’s algorithm is different – it uses live data, including possible pick up and drop off sequences, updating its calculations constantly.

From there, it starts by eliminating the most inefficient routes, as well as those falling outside any constraints set by the operator (such as geographical ones, wait times, and so on) – in a “pruning the tree” approach. Once that’s done, TRC’s algorithm considers the remaining combinations at a much higher speed.

What's more: our routing engine centralizes and commingles scheduling across modalities: on-demand, shuttles, stop-to-stop, paratransit, pre-booking, and fixed route transit. It integrates all modes of transit — working smarter, not harder, across the entire transit ecosystem.