How data analytics can help predict passenger demand

An increase in understanding among travellers and improvements in traffic management is helping to drive growth in transport predictive and data analytics, says a new report.

The Research and Study forecast says the global transport analytics market is set to reach US$37.4 billion by 2028. An increase in awareness among travellers of how to streamline their own travel, as well as the reduction of congestion and improvement of traffic management, is helping foster this growth.

Supporting this are the transit agencies that have invested heavily in data processing in the last decade. Yet, not only are many still deciding what to do with the data they have captured and how to link it together, they also need to justify the investment.

“Just aggregating and averaging isn’t the right answer,” says Sue Walnut, Product Director UK and Ireland, Vix Technology– a transit solutions provider. “You’ve got to take a lot of other stuff into account like, the day of the week, whether it’s school holidays, whether there are any big events, or weather conditions. All this needs to be taken into account before you can just spit out some information.”

Vix is currently fine tuning the algorithmic logic in its Prediction Engine to help provide additional key information for cities, partners and passengers.

“We’re looking at what else could we bring into our Prediction Engine, what brings value, seeing what other data sets bring value, and frankly, what data sets don’t bring value,” explains Walnut. “It’s really easy to take data and make graphs and generate figures but what’s more challenging is finding out [from the data] something that you can change. Something that makes a difference.”

The Prediction Engine can predict passenger demand based on a variety of factors, such as weather, holidays, and events. By using this data to optimise service delivery, transit agencies can ensure that public transport services are available when and where they are needed most.

“We want to work with our customers to have our systems support their investments,” says Walnut. “A lot of these investments are significant: it doesn’t help us and it doesn’t help them to insist that they reinvest or do everything a completely different way. We work really hard to meet our customers and help them add value between ourselves and their other suppliers.”

Assisting operators

Data analytics involves the use of mathematical and statistical techniques to extract insights from large and complex datasets. By collecting and analysing data, transport agencies can optimise their services, identify areas for improvement, make data-driven decisions, and improve passenger experience.

Predictive analytics, on the other hand, involves using historical data to make predictions about future events. By anticipating future events, transit agencies can be more proactive in their decision-making and improve the overall quality of service.

“The needs of the passenger are the needs of the city,” says Walnut. “Cities have an interest in sustainable cities and modal shift. Operators are also looking for efficiencies to make sure they are continuing to provide the best service to the public at the best possible price.”

Improving the passenger experience

Walnut reveals that although a lot of the company’s “bread and butter” is working on predictions and looking at the data that comes in and what’s happening today, this year it will also delve more into patronage analysis predictions.

This will be particularly helpful for bus services. In the US, most bus companies charge a flat fare which means it is challenging to know where people get off.

“When you have the granularity of knowing when people are getting on and off vehicles you can start asking if that pricing is correct,” she says. “Are there ways in which it could be priced differently to attract people from different areas of the route? So those are typically where that information is really useful for [transit agencies’] commercial teams.”

For passengers it will provide information on how full or empty a train or bus is, helping the enjoyment of their travel experience.

She says: “It can help them make decisions, particularly on a frequent service where they say, ‘Well I’m not going to get that bus because it’s rammed, I’m going to wait for the one after because I can see it’s less busy and only two minutes away.’”

In the UK, Transport for Greater Manchester can now visualise a complete reporting and monitoring data feed, provided by Vix, in near real-time to the agency’s systems. The data is aggregated, consolidated, and filtered to provide daily reporting of fare tap-ins, trips, transactions, and settlements.

Remaining challenges

Walnut warns that while predictive analytics indicate where you could change scheduling or amend a route to optimally run your buses, that doesn’t mean you automatically should.

If passengers can’t work out where or when they need to go to the bus stop then people won’t know where they are. The biggest challenge with predictive analytics isn’t necessarily the data, she says, it’s doing the right thing with the data that’s available through analytics.

Likewise, it will be paramount for transit operators to keep up with the speed of change in such a growing market. Public transport agencies will need to adapt to advancements in technology, changing consumer preferences, and increasing environmental concerns. It will become increasingly important for them to use analytics to stay competitive and meet the evolving needs of their customers.