Self-Driving Cars in Australia

Self-Driving Cars in Australia

Filed under Information Centre

Is Australia ready for driverless cars? Many say it’s the next step towards making our roads safer.

We’ll take a look at how a driverless car works, and whether it’s really safe, as well as the advantages that driverless cars will have for our community.

Driverless Cars in Australia

One of Ford’s 3 global research centres is located here in Victoria, and the company aims to bring a fully functioning self-driving vehicle to market by 2021. They've invested $50 million into Australia in 2017 to expand testing and driver-assist technology testing capabilities.

Vic Roads has also been involved in funding a project to create a level 4 autonomous car, combining Bosch technology with a Tesla Model S to create a highly autonomous vehicle, one of just 5 worldwide in 2016.

In SA, RAA is partnering with Flinders University to trial a self-driving shuttle bus on the Roads around the Tonsley Campus, before moving to other campuses, the Flinders Medical Centre, and eventually public roads in and around Tonsley, Oaklands Park and Marion.

How does a self-driving car make ‘decisions’?

A self-driving car acquires awareness of the environment around it via a variety of sensors that include cameras, radars and LIDARS.

The level 4 autonomous Tesla built in Victoria had 60 additional components, including 6 radars, 6 LIDARs, high-resolution GPS, and a stereo video camera connected to 13 computer networks and linked by 2 km of copper wire.

Autonomous vehicles will make decisions based on their own speed, weather, road conditions, distance, and other data collected by its sensors.

Ideally, the vehicle will be gathering and processing this information fast enough to avoid dangerous circumstances at all, resulting in a much safer journey than when most human drivers control the vehicle.

This is no reflection on the incompetence of people as vehicle operators, but rather just a result of the advanced technology that guides these cars.

Navigating tricky ethical questions

The biggest source of apprehension around self-driving cars comes when the software programmer has to decide who the car will protect in the event of a potential accident.

There’s a notable reluctance as consumer’s reveal they’d be unlikely to purchase a car that didn’t put the occupant's safety first, but they would prefer other people to buy vehicles that considered the ‘greater good’ the harm minimisation programming activated in a potentially dangerous situation.

If a child were to run onto the road in front of the car, where there’s a steep drop-off from the side of the road, should the car continue on its path and risk hitting the child, or swerve to avoid the child, risking the lives of its occupants?

In Singapore, testing of autonomous vehicles in public parks shows that the majority of accidents and near misses occur not because of a malfunction, but when people jump in front of the vehicle to ‘test’ to see if it will stop in time.

Whenever pedestrians and cyclists share the road with cars, there’s going to be risks, including risks of behavioural patterns that can’t be foreseen.

To get involved in the discussion and have your thoughts considered in the global development of autonomous vehicles, you can submit your thoughts and ideas here at the Moral Machine. Share your results, and get others involved in the discussion too.

Removing human error decreases accidents

Ford’s driverless car testing in Australia has shown that self-driving cars have fewer collisions than human-operated vehicles.

In fact, 90% of all accidents are caused by driver distraction. This is where the self-driving car steps in - machines don’t get distracted.

They don’t drink-drive, get overtired, make phone calls while driving, or turn around to talk to the kids in the back seat.

Those at the front-line of self-learning software admit that new types of accidents will arise, but are confident that as sensor technology improves and self-learning software gets smarter the number of incidents will continue to decrease.

For example, combining GPS data with visual clues, the Oxbotica can now safely navigate tunnels and areas where tree cover might interfere with a strong GPS satellite signal.

Google's fleet of self-driving cars has been active since 2009, driving 804, 672 kilometres without a crash. Compare this to human drivers, who get into at least one accident every 804, 672 kilometres.

"If the car is in autonomous mode and the car is in an accident, Volvo is responsible for that accident."

Volvo is confident about its autonomous driving technology and has committed to take responsibility for any accident that occurs when the car is in autonomous mode at the Federal Government enquiry into social issues around driverless cars.

They did set limits on their responsibility to manufacturing defects, so if poor maintenance such as bald tyres contributes to the crash, the driver will have to take some responsibility.

Autonomous controls that are already in place in our cars

The introduction of self-driving cars isn't as scary as it appears when you consider the override technologies that have already been built into our modern cars to 'assist' negligent drivers to avoid accidents or reduce harm.

Starting with anti-lock brakes, which override the driver's application of their foot to the brake pedal to produce smoother, safer braking, and sensors which provide rear camera views and warn of nearby objects.

Many cars now have adaptive cruise control, hands-free self-parking, lane keep assist, lane departure warning and driver fatigue warnings - just a few of the many ways that humans have delegated our safety to technology rather than our own abilities.

How will society change?

Self-driving cars will give access to practical, flexible transport to people who now have to rely on taxis, family members and public transport. This will give back independence to anyone who can’t drive due to loss of vision, age, or disability.

For the rest of us, it could potentially improve our quality of life. Driving can be tiring, difficult and dangerous. It’s also a big waste of valuable human time.

Families that now have 2 cars and a double garage would be able to use one car or shared ride, as the car was taken to work in the morning could drive itself back to take kids to school, pick them up at the end of the day, and other tasks during the day. The double garage could be used as extra space for the family, opening up the home to renovation possibilities.

Imagine if instead of spending hours stuck in traffic, getting frustrated and putting your life in the hands of other drivers just as frustrated, stressed and tired, you could use the time to do whatever you pleased?

Productivity would undoubtedly improve. Anxiety and stress would be reduced. Self-driving cars will lead to steadier traffic flow, less acceleration, reducing emissions as well as creating a smoother journey for the vehicle’s occupants. It’s a win-win.

Reduction in personal car ownership

It’s also predicted that personal ownership of vehicles will decline, with an international KPMG survey revealing that 59% of car manufacturing executives anticipate that more than half of all car owners today will no longer want to own a car in 2025.

Driving tests will change too, so you might be ‘licenced’ to ‘drive’ a vehicle with a certain type of automation system.

How will car manufacturers make money?

It’s predicted that car manufacturers will shift their focus from selling directly to the consumer, instead of building driverless, electric vehicles aimed at tech firms like Google and Apple.

Private transport becomes an intimate pay-to-use service and money is made via the peripheral digital services that accompany these ‘robot-taxis’ of the future, rather than relying on revenue generated primarily through car sales.

The car will become an integral part of a modern, connected lifestyle, the true realisation of the internet of things. For example, Bosch plans to update self-driving trucks with real-time data on the route, congestion, detours and unloading facilities at the destination, allowing the driver to get other things done while the truck drives itself.

It is also anticipated that there will still be sales of cars through dealerships in the traditional way and that manufacturers will maintain relationships with vehicle owners through the ‘connectivity’ used to update the car with the latest software - much like Tesla do right now.

Could hacker’s wreak havoc on autonomous vehicles?

An autonomous car would certainly be more vulnerable to hackers than a traditional car, with more control units, greater computing power, lines of code and wireless connections to the outside world. There’s potential a hacker could exploit the privacy of the owner, including financial information, or cause the vehicle to fail to start or to crash.

Major advances in software security will ensure that critical systems, especially interfaces connecting the vehicle to the outside world, have strong fundamental security. Systems crucial for safety will be separate from car and infotainment systems, and from the networks that facilitate secure communication between control units. There can be over 150 networks in an autonomous vehicle.

In addition to a careful, thoughtful initial setup, any arising vulnerabilities can be patched seamlessly with ‘over-the-air’ software updates in real-time.

When will we see autonomous cars on the road?

Autonomous vehicles with Level 2 and 3 automation are already being used in mining in Australia. With what’s referred to as ‘conditional’ or ‘partial’ automation, these trucks are monitored by human drivers.

We’ll likely see shared ownership of autonomous vehicles and self-driving public communal transport vehicles as a ‘pay-to-use’ shuttle first (Routemaster bus). Right now the costs of hardware such as sensors mean the initial costs are still too high.

The predicted benefits of autonomous vehicles being a 90% reduction in accidents, 40% less congestion, up to 80% fewer emissions, and 50% parking space saved are so substantial that development and adoption of autonomous technology will continue.

In Australia, we’ll need to see changes to the Australian Road Rules, which assumes a driver is human rather than a ‘system’ before self-driving cars are used outside of approved tests.

Who are the manufacturers to watch?

There are at least 33 tech and auto corporations internationally involved in the development of AI for autonomous vehicles.

It's predicted that Nissan will be first to market in around 2020 with a semi-autonomous Nissan Leaf.

Other key players are Audi, with the A6 Avant; Mercedes Benz with the S500, and BMW X5, all emerging in 2013/14.

BMW also released a concept car in June 2016 called the Mini Vision Next 100 which changes colour based on user preferences.

BMW's vision is an expansion of the sharing economy, where any car of the fleet is shared by many people. Driving autonomously to the pickup location, the car updates its settings to match your personal preferences.

Others include Ford Smart Mobility, Google, Tesla, Apple, Honda, Intel, Uber and Volkswagen, a Chinese bus manufacturer Yutong ... there's plenty of competition to bring an autonomous vehicle to market.

Quick Quote

By clicking Get My Quote, I acknowledge and accept the Terms & Conditions.

Did you enjoy this article?

We can deliver articles like this one straight to your inbox. Simply enter your email address in the box below and we'll take care of bringing you the best content.

Email Sent!

Please enter a valid email.

We make finance a Positive experience