Cars are a huge part of our lives, but they also have their fair share of challenges, like safety and the need for driving skills. To eliminate some of these challenges, automakers and tech companies are embracing modern technologies, like self-driving cars. Though the idea of self-driving cars has not materialised as fast as we thought, most leading automakers have made prototypes that are already on our roads for tests. For example, on October 8, 2020, Waymo introduced its first fully autonomous vehicle on public roads and invited members of the public to experience their rides. Therefore, we are rapidly moving from seeing these cars in science fiction to owning them. Read on to learn more about the current status of self-driving cars as well as their future.
Table of contents
What Does ‘Self-Driving’ Mean?
According to the Society of Automotive Engineering (SAE), there are six levels of autonomy, starting from no assist (Level 0) to Level 5 (Full autonomy). Here are the levels of automation:
- Level 0: These are vehicles with no driver assistance technologies.
- Level 1: Vehicles with an advanced driver assistance system that assists the driver with steering, braking or acceleration, though not simultaneously. Such systems include features like vibrating driver’s seat to alert drivers whenever they drift out of their lane and rear cameras.
- Level 2: These are vehicles with an advanced driver assistance system that can either accelerate or brake simultaneously without the driver’s actions. The driver should remain alert behind the wheel to act when necessary.
- Level 3: These vehicles come with an automated driving system that can perform certain driving tasks, like parking, without the driver’s actions. Level 3 vehicles can make their own decisions, but the driver must remain alert and take control whenever the system fails.
- Level 4: These vehicles can perform all driving tasks in certain circumstances. In such circumstances, the human driver doesn’t need to pay attention.
- Level 5: In these vehicles, the automated driving system does all the driving in all circumstances. This means that level 5 autonomous cars can go anywhere at any time. They also don’t need a human occupant on the driver seat as there is no need for human intervention.
How Self-Driving Cars Work
Self-driving vehicles are powered by AI technology. The developers of autonomous cars use huge amounts of data from image recognition systems and neural networks, and machine learning to build systems that can drive without human intervention.
After identifying the patterns in data through neural networks, it is fed to the machine learning algorithms. This data helps the neural networks recognise pedestrians, street lights, curbs, trees, streets and other items in the environment. The car will most like use systems like cameras and sensors to generate the data, meaning the system drives, the more data you add into the deep learning algorithms, helping the car make better decisions.
The Current Situation with Self-Driving Cars
Though cars cannot drive themselves yet, most recent models have self-driving features. Some of these features include:
- Lane-centring steering: This feature helps drive to remain in their lane by automatically nudging the vehicle towards the lane marking on the opposite lane mark every time the driver crosses the lane markings.
- Hands-free steering: This feature centres the car when the driver is not holding the steering wheel. However, the driver should not confuse it with full autonomy.
- Adaptive cruise control: This systems automatically maintains a pre-set distance between a car and the vehicle in front.
Today, most cars in the market contain Level 1 autonomy, but some offer Level 2 autonomy, like GM’s Super Cruise and Tesla’s Autopilot. When using a vehicle with Level 2 technology, the driver can let the car manage the speed and steering, but they should remain alert to seize control when needed.
Other manufacturers, like Honda, are focusing on Level 3 autonomous systems that will enable the car to fully drive itself in less complicated conditions like low-speed in traffic. On the other hand, Google’s Waymo and ford have skipped levels 2 and 3 to focus on Level 4 technology — Full autonomy within a geofenced perimeter. Currently, level 4 and level 5 self-driving cars are not commercially available.
The projection on when Level 4 and Level 5 vehicles will be commercially available may vary among manufacturers. For example, according to Elon Musk, Tesla will have Level 5 autonomy by the end of 2021. Also, both Volvo and Ford project to release a Level 4 vehicle for commercial use in 2021. On the other hand, Google’s Waymo promised to be available for public use by 2020, but that was not achieved.
Even if self-driving cars are not available for public use, they are already in use in other ways. For example, in 2019, Lyft partnered with Waymo to provide a fully autonomous ride-sharing service in the Metro Phoenix area. There are also autonomous street-sweeping vehicles already working in Hunan province, China.
Why Has Self-Driving Taken Too Long to Arrive?
Self-driving technology still has a long way to go before vehicles become truly autonomous. Though artificial intelligence has enabled numerous breakthroughs in various fields, it has not mastered driving as quickly as developers predicted. Since safe driving entails more than just operating the vehicle’s controls, like predicting human behaviour, it has proved to be trickier than anticipated.
When it comes to the hardware, the self-driving systems rely on numerous sensors and high-speed cameras that must improve and become cost-effective to enable mass production. Also, numerous regulation issues need to be solved before fully autonomous are allowed to use public roads. For example, there is no exact definition of self-driving as well as guidelines to ethical and liability concerns, like who should bear the fault if an autonomous vehicle causes an accident. In addition, manufacturers need to convince people that self-driving is safe.
Pros and Cons of Self-Driving Cars
Below are some of the benefits of self-driving cars.
Prevention of Accidents
The number one benefit of an autonomous vehicle is safety. Most accidents worldwide are due to human errors and poor choices like drunk driving, and since autonomous vehicles can eliminate these risk factors from the equation, they may be safer than human-driven cars.
The reduction in accidents comes with numerous economic benefits. For example, if there are fewer accidents, there will be less strain on the healthcare system.
The ability to communicate with each other is a major benefit of self-driving cars. Thanks to the ability to communicate in real-time, cars can maintain an optimised distance from each other and determine the best route to avoid traffic jams.
Since most self-driving vehicles’ manufacturers embrace e-mobility, they will significantly reduce emissions.
Despite having all the above advantages, self-driving vehicles have some disadvantages, including:
Since automated cars will be sharing the same network protocol to talk and coordinate with each other, there is a potential risk of hacking. Even a small hack into such systems in a busy road can cause significant damages.
With the introduction of self-driving cars, people who earn a living through driving will lose their jobs. There are millions of people employed as truck drivers, taxi drivers, bus drivers and chauffeurs.
Even if self-driving vehicles provide a significant societal cost-saving in the long term, the initial cost of acquiring them during the initial stages will be high.
While there is a high possibility of preventing accidents using self driving vehicles, the technology does not completely eliminate the risk of accidents. If any part of the vehicle or the software fails, the vehicle could put the occupants in more danger than if a human driver controlled it.
From the above analysis, it is evident that there is plenty of debate about self-driving cars. Everyone agrees that autonomous vehicles are the future, but they also have some serious concerns. Some strides have been taken towards full automation, but there are numerous roadblocks along the way. Contact us for more information.