The pursuit of creating fully ‘driverless’ cars has made way for some of the biggest technological advances ever seen in the automotive industry.
While total autonomy still remains a thing of the future, companies are now predicting that a self-driving vehicle could become a reality as early as 2021.
Accelerated sensor, camera and data technology has meant that companies such as Google, Tesla and Volkswagen have been able to test or at least develop vehicles that have increased levels of autonomy.
But, as it stands, there is still not one completely driverless car on the market.
With rafts of legal implications and safety issues circling this new automotive technology, we take a look at what driverless actually means in today’s world and what we can expect for the future.
To qualify as fully autonomous (also known as 'driverless' or 'self-driving'), the car must be able to get from one destination to another using just a computer, navigating and manoeuvring through roads that have not been adapted for its use and in a variety of different conditions.
If any form of human control or intervention is needed, the car will not be classed as fully autonomous.
This level of autonomy means most functions will still be controlled by the driver, but one specific function - such as steering or braking (but never both) – can be performed automatically.
This level has been around for several years and includes cruise control and lane assistance.
This is where our current technology is today. Computers can take over multiple functions from the driver, including steering and speed in unison.
The driver does still need to be engaged, monitoring the surroundings and ready to take over control of those functions at any given time. Tesla’s Autopilot is at level 2 autonomy, using functions such as traffic-aware cruise control.
This type of autonomy means the vehicle is able to cover nearly all aspects of driving. This includes monitoring surroundings, braking and controlling the steering – but only in certain conditions.
The driver is not required to monitor the situation, as is required with lower levels, but must still be on hand to intervene if needed. The new Audi A8 is the first production car to have level 3 autonomy.
Level 4 automated cars can drive themselves, although with a human on board. The car can start, drive, monitor surroundings and park in a larger variety of environments – but not all. In certain conditions, the driver can switch to fully autonomous and relax.
If the car becomes confused by a situation, it can request help from the driver but will also continue to drive if the request is ignored. There are no production cars currently at this level.
A car that reaches level 5 will no longer need any human control. It will be able to complete all driving tasks, under any conditions and understand all scenarios including traffic jams.
These cars would no longer require pedals or a steering wheel.
There will be several complex features that have to work in unison to create a successful self-driving car.
It will include a variety of sensors and cameras that will monitor other vehicles on the road and detect objects, traffic lights, road signs, curbs and lane markings. For example, Lidar sensors use laser beams to create a 360-degree image of the car's surroundings.
An effective GPS system will be vital in ensuring the car knows how to get to a pre-determined destination, while an intricate on-board computer will then use all of the data collected from the surrounding features to instantaneously manipulate the car in the correct way.
Every manufacturer will have different types of technology, especially in the early testing stages – there isn’t one system that fits all.
Cars today are becoming increasingly advanced – even the most basic cars now come with a lengthy list of options to consider. But what do you actually need to spend your money on? Our navigates you through the extras worth considering.
There have been a number of accidents involving autonomous cars, including one that resulted in a fatality. Some of these were caused by issues with the capability of the car.
For instance, autonomous cars must learn to identify objects in their path, whether it's a piece of litter or a pedestrian. It must also know when there is a lane closure, a difference in the width of the road or even where to position itself if an emergency vehicle needs to pass.
The car then has to decide instantly whether to swerve, brake or continue.
In some cases, semi-autonomous cars have not been able to make the correct decision and it has resulted in accidents. This is an ongoing problem that manufacturers will have to find a solution for before a car can become fully driverless.
The driver can also cause safety implications too. We have yet to see a fully autonomous car, so the driver of the vehicle is still responsible for its actions – regardless of whether they are in autopilot mode or not. So, reading a book or simply not paying enough attention to the road can pose a risk just like it would for traditional cars.
This reinforces the need for drivers to be fully aware of the limitations of the car to ensure the safety of themselves and others.
All of the above also reiterates the need for effective and thorough testing of cars before they are available to buy.
However, when the technology is advanced enough for a fully driverless experience, this could mean that our roads actually become safer. Using computer software could eliminate human error such as carelessness, tiredness and drink driving, resulting in potentially safer roads.
The morality of autonomous cars is one of the biggest issues that computer coders and manufacturers will have to tackle. The cars will have to be hard-coded to choose one action over another in varying scenarios – but in the worst circumstances both options might be tragic.
For example, if a car had to swerve a motorcyclist that’s heading in its direction at high speeds, but in doing so could hit a group of teenagers stood at the side of the road. The car would have to make the decision to potentially hurt one over the other – and all within a split second.
These moral dilemmas will have to be acknowledged and a course of an action for all the varying scenarios will have to be programmed into the future self-driving car.
Lawmakers will also need to decide how and who will be held accountable if these types of accidents occur.
The move to launch semi-autonomous cars on the roads of the UK is well under way. But aside from creating the suitable technology, testing in the ‘real world’ also has to be completed before these cars can be put into production.
The Department of Transport has published a code of practice that provides a clear set of rules for the manufacturers wanting to test cars. If compiled with, this should allow firms to begin ‘real-world testing of automated technologies’ in different locations around the UK.
‘Real world’ means driverless cars can be tested on public roads under certain conditions.
Some of the Department of Transport recommendations include:
Laws have not yet changed to reflect this growing technology. As it stands, the driver is still required to be in charge in all semi-autonomous vehicles, so the same laws apply for these cars as they do to traditional cars.
However, the government began a three-year law review in early 2018 to ensure reasonable regulations are in place before self-driving cars are on UK roads. Here are some key areas the review will focus on:
Insurance cover will undoubtedly have to make lots of adjustments to fit the new wave of driverless cars expected to arrive in the next ten years.
Here are ways traditional car insurance cover could be affected:
Here’s a overview of some of the most up-to-date tech available today.
Even though we haven’t yet seen a car that reaches level 5 autonomy (a car that no longer needs human intervention of any kind), many manufacturers have created models that include impressive semi-autonomous features.
has become a well-known name because of its headway with autonomous technology. However, the Autopilot capabilities are still at level 2 and are synonymous with semi-autonomy: the car can take over one specific function such as steering but the driver still needs to be engaged and ready to take over.
The most recent hardware is called Enhanced Autopilot. It's essentially a driving-assistance feature. It can auto-steer, keep you in your lane, brake and accelerate the car forward, as long as the driver is still paying full attention with hands on the wheel. The feature can also self-park.
Like Tesla’s Autopilot, Volvo’s pilot assist is meant to aid the driver rather than provide fully autonomous travel. Again, the driver always needs to pay attention and keep their hands on the steering wheel, while the car can help to steer, keep you in the correct lane and maintain a safe distance from the car in front.
Depending on the road, the assistance can switch on and off so the driver should never solely rely on the technology. If your hands come off the wheel the car will sound an alarm.
A further addition to the adaptive cruise control is the Stop & Go feature. This warns the driver of a car sharply breaking ahead with a visual and then an acoustic signal, plus it will automatically depress the brakes if no action is taken.
CoPilot Assist sits at level 2 autonomy, which means it can take over one function automatically but the driver still has to be aware of the surroundings and ready to take over at any point. It also includes adaptive cruise control, lane centring assist and active park assist.
Just like Ford, BMW and Volvo, ProPilot will adjust the speed to suit the traffic ahead and keep the car centred in the lane. It will bring the car to stop if it deems it necessary and can automatically parallel and bay park once the driver selects a proposed system.
The new A8 from is the first production car that can reach level 3 autonomy – this type of autonomy means the vehicle is able to cover nearly all aspects of driving, but only in certain conditions. It doesn’t mean you can sit in the back and take a nap while the car drives you to your destination, but it does allow the driver to remove their hands from the steering wheel for longer periods of time while the car takes over during stop-start traffic.
Once the car hits speeds over 37mph, the driver has to take over.
This a big step in the industry but the law isn’t quite there yet. Until legislation changes, the level 3 features aren’t allowed to be put into use. Testing and alterations to suit each country’s legal stance will determine if and when its full capabilities can be put into practice.