CIO

Transurban takes express route to smart roads and driverless vehicles

Company's sensor rich roads keep congestion down and enable an autonomous future

The price you pay for not using the tolled express lane on Washington DC’s Interstate 95 highway is frustratingly apparent.

"In the express lane you will literally drive past the general purpose lanes where everybody's backed up in traffic and you get to see the value proposition,” explains Robert Dean, head of technology for Transurban – the I95 express lane’s owner and operator in North America.

Dean describes paying the toll for the I95 express lane – which is due to be extended following a deal signed last week – as the equivalent of a “guaranteed time to delivery” for your journey.

"If you take general purpose lanes you might be able to get there on time,” he told CIO Australia at the AWS Re:Invent conference in Las Vegas last month. “But a lot of times you're not going to get there. We take the express lanes because we – for all intents and purposes – guarantee a certain speed on those lanes at a minimum.”

Although the advantage of taking the tolled route is less obvious in Australia, where Transurban wholly or partly owns 13 roads, it is hard to dispute. According to data from satnav maker TomTom, drivers in Sydney, Melbourne and Brisbane save more than 310,000 hours of travel-time each workday on average by taking a Transurban toll route.

Nobody enjoys paying road tolls, so for Transurban’s value proposition to be sustainable, its routes must remain congestion free.

But with the number of cars on Australian road steadily rising (up 2.1 per cent to 18.8 million at last count) and their use during peak hours increasing too (around two in three Australians drive to work, up by half a million since 2011), keeping toll roads flowing freely gets more challenging every day.

Transurban is rolling out IP-enabled technology along its fibre-lined roads to gather data, managing lanes from control rooms that crunch data from driver mobiles, video cameras, weather sensors, government agencies and demographic studies. The company is also preparing for the unknown effects autonomous vehicles will have on the road network, and running a series of trials in Australia. At present 40 per cent of our Transurban’s workforce is in technology with a daily focus on data, network performance and security.

“The better that you can become at predicting as opposed to reacting, allows you to manage that road more effectively,” Dean adds.

Life in the fast lane

Transurban’s network in Australia is vast, the company wholly or partly owning the Cross City Tunnel, Eastern Distributor, Lane Cover Tunnel, and a stretch of the M2 in Sydney, Melbourne’s CityLink and controversial West Gate Tunnel, and Brisbane’s Clem7, Go Between Bridge, Logan Motorway and Airport Link M7 among others.

An artist's impression of the Melbourne West Gate Tunnel entrance
An artist's impression of the Melbourne West Gate Tunnel entrance

Along its routes, the company has laid 700 kilometres of optic fibres which are connected to 100,000 pieces of technology and safety systems.

As well as the antennae that read the e-tags of vehicles passing under the gantries, there are high definition cameras to read numberplates, dynamic speed signs, and intrusion detection to pick up vehicles travelling the wrong way down a lane. In tunnels there are air meters, exhaustion fans, smoke detectors and safety systems.

“We always say we have the most sensor rich roads in the world,” says Dean.

Increasingly, the sensors and devices are being updated so they are IP-enabled, so they can be more easily read and operated from a central control room.

“Before they were just black boxes, not connected to anything” explains Mithran Naiker, general manager, infrastructure and service delivery at Transurban. “All those devices were getting managed by people at the road, now we’re trying to centrally manage it, centrally deploy it, make it highly automated so no one on the roadside has to go and touch anything besides plugging a box in.”

The company is currently exploring edge computing, so that for example a camera on a gantry has the processing power to read a numberplate rather than a human in a control room.

The devices are business critical. Without the e-tag readers and cameras, revenue is at risk, and if sight or safety systems are lost, the road must close.

“If we can't see what's happening on a road and we can't control the safety systems we need to shut it down,” Naiker adds.

That means the devices chosen are robust, with lots of redundancy built into the system.

“You have to almost be bulletproof,” says Dean. “That doesn’t mean something never fails, but that its ok if something fails if something else can pick it up.”

Page Break

For customers, the device rich roads have meant Transurban has been able to roll-out an app for drivers who don’t have an e-tag. LinktGO is a GPS-enabled mobile app that allows the 50 per cent of toll-road users who typically use Transurban roads fewer than four times a year to see their toll travel in real time and pay trip-by-trip using their smartphones, with no ongoing commitment.

The app, the company’s websites, its Adobe Experience Manager Content Management System, consumer mobile channels now reside within Amazon Web Services, linked by API gateways, part of Transurban’s cloud-first strategy which is increasingly incorporating operational data like workloads from its legacy tolling platform as well.

“It ramped up really quickly,” Naiker says. “Now there has to be a reason why we wouldn’t put a workload in the cloud.”

The cloud move also helps remove some of the risk around building new roads, Dean adds.

"You can actually spin up the technology that supports that road very simply, by literally pushing a button and spinning up a new instance in AWS, versus having to buy all the kit, install all the software, get it all configured, get a data room," he says. "Because usually the technology always comes after the road, that's always a nervous moment – can you actually toll? – and this allows you to do a lot of that ahead of time. It's a huge risk reduction."

Information highway

Real-time speed, volume, vehicle occupancy, travel time, video, tag, weight in motion, GPS, driver mobile and weather data is crunched by Transurban’s in-house, 35-strong traffic modelling team to keep journey times to a minimum on the company’s roads. In response they can change speed restrictions, reverse lane direction and issue queue warnings to help ease traffic build up.

Artist's impression of the proposed bridge over Maribyrnong River in Melbourne
Artist's impression of the proposed bridge over Maribyrnong River in Melbourne

Such systems installed on Melbourne’s CityLink as part of the Monash Freeway Upgrade have seen average speeds increase by 40 per cent at peak times, the company says, allowing for 20 per cent more cars to travel in each lane.

There is a safety benefit too. The serious injury rate per 100 million vehicle km travelled on Transurban’s roads is estimated to be up to 80 per cent below comparable state averages on the broader network, according to the company’s own calculations.

On construction projects, the real-time data and additional sources like land use, employment drivers, customer behaviour, socio-demographic breakdowns and mode-share patterns are used to “understand network friction points and determine the optimal time to deploy capital” in other words, which construction should be done first to best avoid delays.

The rise of driverless and semi-autonomous vehicles, presents even more opportunity for data collection and exchange, Dean says. Transurban has established itself as a lead in local trials.

In August last year the company and the Victorian Government began a two-year trial of how automated vehicles interact with road infrastructure on the Monash-CityLink-Tullamarine corridor. The first phase of the trial, which featured connected and automated vehicles from BMW, Mercedes, Tesla and Volvo tested how car features like lane keep assist, adaptive cruise control and traffic sign recognition respond to tunnels, road works, electronic speed signs and congestion.

In the US the company is about to pilot an app which locates where workers are on a stretch of road.

“At some point you can feed that into the autonomous car,” Dean says. “So I can tell your car there's a worker over the next hill, which will allow the car to move lanes sooner which is good for safety. It's also good for traffic because as you know if we all move over at the very end we bottleneck.

“The road infrastructure and bandwidth means we can share data with the [autonomous] cars, with the hope that the car will then share data with us as its progressing down the road,” Dean says.

Potentially, Dean adds, there could be special lanes for electric vehicles to charge along the highway.

“There’s a lot of opportunity there,” he says.