Getting personal on PT

| March 15, 2020

“I think public transport is painful. It sucks. Why do you want to get on something with a lot of other people, that doesn’t leave where you want it to leave, doesn’t start where you want it to start, doesn’t end where you want it to end? And it doesn’t go all the time … That’s why people like individualised transport that goes where you want, when you want.” Tesla founder Elon Musk

Electric cars aren’t the answer to Melbourne’s transport problems, Mr Musk. They might cut carbon emissions, but they won’t reduce our traffic, which is bad enough already.

But the billionaire has a point. Nearly 70 per cent of trips in Melbourne are made by car – because trains and trams don’t service every suburb, or maybe the tram comes but it’s slow, while the buses are infrequent. And cross-city trips are hard to coordinate, especially on the weekend.

Maximising efficiency

Monash IT specialist Professor Mark Wallace is working on an answer, which he calls “personalised public transport”. The idea is to help Melbourne commuters make more efficient use of the existing road and transport network by tailoring individual journeys. Adding more trains, trams and buses to the mix would help, of course, but in the meantime, Professor Wallace and his fellow researchers are trying to improve how people travel now.

“If you’re going to cut down congestion, you’ve got to have 40 people in one vehicle (such as a bus) instead of one person each in 40 vehicles,” Professor Wallace says.

“So the larger those vehicles, where there’s shared transport, the better for congestion.  But if we can get more people on to the larger vehicles such as buses and trains, by having a really efficient connection between the home and the station – even if it means people are driving to the station – then you’re still attacking congestion.”

Professor Wallace’s particular research interest is modelling and optimisation. He and his research team are investigating data on trips taken by drivers around the city, to see if adjusting the route or the time of departure would mean a quicker journey.

The project aims to use optimisation and simulation to see how congestion could be reduced if a proportion of travellers followed the routes suggested to them – by car, bus, train or tram, or a combination of them. One of the questions of interest to the researchers is “buy-in”. How many drivers would need to use the service for it to make a difference to the city’s congestion? And how much of a difference would it be?

The problem is complicated. As Professor Wallace points out: “if you’ve got 10,000 vehicles, all looking for the shortest path, then it won’t be the shortest in time if they all end up using the same route.”  And that’s just one of the many issues the researchers are tackling, he says.

Managing multiple alternative routes so that they all achieve “the minimum overall time, and how to do that in a way that’s fair, is a serious technical problem”, he says.

Playing a bigger game

Team member Daniel Harabor developed an algorithm, now widely used in computer games, that will be enhanced as part of the research project.

“In computer games, you have all these objects rushing around the screen, and you have to coordinate them, and you have to keep them all going in real time,”  Professor Wallace explains. But the Melbourne traffic grid is much larger and more complex than any game. For a game strategy to be adapted to the transport network, it has to be “scaleable”.

The three-year project is drawing on already-gathered information about driver journeys to predict road and public transport use. “We’re using what data we can, historical data for Melbourne, and we’re saying, ‘Right, supposing we were routing all that data, what difference could we make?’”

Professor Wallace says. “We’ve bought some hardware, and we’re going to simulate arrivals, or requests coming in, tens of thousands of requests, and routing them all.”

Enhancing connections

In a separate project with the Department of Transport, researchers are also examining whether existing bus and train services could connect better – what changes would need to be made for the bus to wait for a train, for instance, or vice-versa?

These protocols are commonplace in European cities but still remain, as Professor Wallace says, “a novel concept” in Melbourne, where the population is spread more thinly over a larger area.

“What stops some people from taking public transport is sheer prejudice – they don’t want to sit with other people – but for many others it’s the uncertainty and the additional time required,” he says.

“If our research means that public transport takes only a third longer than a car, I think quite a lot of people would take it, but if it takes twice as long as in a car, then there’s going to be a smaller take-up.” So if someone spends 30 minutes driving to work, they could be prepared to spend 40 minutes on a train and bus, but reluctant to spend an hour.

The third aspect of the project is to increase the number of people who share a vehicle, “again by coordinating people’s schedules and bringing together a route that can carry the most people with the least extra time involved for each person”.

Past research has shown that one barrier to car-pooling is people’s reluctance to talk to strangers in the morning. Professor Wallace nods in sympathy.

“I think any change in behaviour is hard to encourage; people get very stuck in what they’re used to,” he says. “So any change needs some motivation, if you like, some carrots and sticks, which government could easily propose, so that the way to reduce traffic found in Europe a lot is to make parking very expensive.”

Professor Wallace agrees that “personalised public transport” sounds like a contradiction in terms. No comparable system exists anywhere else, he says. “I think we will come out with some quite dramatic new results in the area of multi-modal transport, and of coordinated routing,” he says. “I’m quite excited from a research point of view.”

This article was published by Lens.