A guide to autonomous vehicles

Autonomous vehicles are gaining traction faster than you think. Here is everything you need to know about driverless cars and the road we’re heading down

Add bookmark

Adam Muspratt
06/19/2018

Autonomous vehicles: An introduction

Vehicles have improved considerably over the last 50 years with an endless list of incremental improvements. We have seen the invention of the hybrid/electric car for unmatched fuel economy, we have seen the introduction of electronic stability control (ESC) for improved safety, not to mention the smart key which has done away with the inconvenience of fumbling around in your pocket for a traditional turning key.

The next big improvement in vehicle technology is a significant technological milestone and game changer.

Enter the autonomous vehicle.

No longer a distant idea found only in the realms of science fiction, the technology is already here and driverless cars are navigating streets today. Over the last five years, the technology has rapidly progressed and everyone is looking to get their hands on the wheel, from existing car manufactures to tech start-ups.

What is an autonomous vehicle?

Sometimes called driverless cars, self-driving cars or automated vehicles, this form of technology refers to a vehicle that travels between destinations without a human behind the wheel. Autonomous vehicles utilize a mix of artificial intelligence (AI), machine learning (ML), radar, cameras, sensors, data, and more, to navigate roads using nothing more than complex algorithms.

SEE ALSO: The automated driving functions with most potential

How does an autonomous vehicle work?

So, what goes on in an autonomus vehicle that gives it the ability to dynamically react to obstacles and make human-like descisions?

There a wide array of autonomous vehicles being developed by manufacturers all over the world, from automobile manufacturers like Nissan to tech giants like Google. Even Uber is getting in on the autonomous vehicle action.

There are also different levels of autonomy (more on that below) and, at different levels, different technologies will be applied.

As such, the internal workings of each autonomous vehicle will vary from manufacturer to manufacturer, but all of them will share the same core components such as synchronized sensors.

There are key components that comprise an autonomous vehicle:

(Source: Shutterstock)

1. GPS Navigation – GPS (global positioning navigation) has been around for decades, giving human drivers the ability to navigate the roads without a map. Similar in concept to applications like Google Maps, a GPS system achieves three things: it sets out the objective; it informs where the car is in relation to the destination, and it also informs what routes are available to get to the intended destination.

2. Radar Sensors – This gives the autonomous the ability to sense how far the vehicle is from obstacles, such as other vehicles. Typically, these are mounted on the front and rear bumpers.

3. Cameras – The technology will vary, but all autonomous vehicles will need cameras to identify road markings and traffic lights. In unison with the sensors, the autonomous vehicle will be able to determine what the obstacles are, the dimensions of objects and objects in the peripheral.

4. Lidar Unit – A Lidar (light detection and ranging) unit gives the vehicle a 360-degree view of its surroundings. A Lidar unit works like a long-range sensor with of up to 100 meters. Typically, it is mounted on top of a vehicle and like a siren, it constantly spins around and beams out lasers in all directions to scan the surroundings and build up a 3D image of the area. It figures out where obstacles are as the lasers hit and bounce off surfaces.

5. Computer/Processor – The computer will analyze all of the incoming data from the sensors, compare it to previous data, and calculate the appropriate action. Some autonomous vehicles will have more than one processor depending on the number of cameras, sensors and navigation tools that the car is housing.

SEE ALSO: Autonomous vehicles: What are the roadblocks? Levels of automation

Experts in the autonomous vehicle industry refer to a scale which categorizes the autonomy levels of different vehicles.

Level 0 – This kind of vehicle has all of its systems operated by a human, such as the steering, acceleration and breaking. It is not an autonomous vehicle.

Level 1 – Vehicles in this category have a range of systems that can assist the driver. Examples of this include cruise control, and autonomous emergency braking (AEB) that utilizes sensors to reduce the speed of a vehicle before it hits something. As a whole, the driver still performs all of the major functions that are associated with driving a vehicle.

Level 2 – Most modern vehicles fall into this category. Level 2 cars allow partial automation and allow the driver to disengage from one or two major functions while the vehicle takes over. Examples of this include systems that manage your speed and steering in situational conditions, such as cruise control and auto-steer.

These level two systems can lock your speed, match your speed with vehicles in front of you and follow the curves of the road you are driving. Otherwise, the driver has to be alert at all time to take control of the vehicle, as responsibilities such as monitoring traffic and general surroundings are not handled by the autonomous vehicle. In addition, cruise control and auto-steer aren’t able to navigate tight bends and residential streets.

Level 3 – At this stage, there is a major technological leap. This kind of autonomous vehicle can actively monitor the environment and take over major functions such as steering, breaking, and land changing – under certain conditions. In addition, the driver is still expected to take over control when the situation changes.

Examples of this include the Audi A8 which is due for release in 2018 and is considered the first level 3 autonomy vehicle. At the press of a button, the Audi A8 can manage the steering, acceleration and breaks in slow moving traffic up to 60kmh/h. When traffic clears up and the conditions for Level 3 autonomy no longer exist, the driver will be alerted to take back control of the vehicle.

Level 4 – At this level, the vehicle will be able to handle all major functions including monitoring surroundings, but often have a human on board. These cars are highly autonomous and truly self-driving, but when it encounters something it is unprepared for it will request that the driver takes over. Situations that the autonomous vehicle isn’t prepared for can include poor weather and off-road driving.

That said, if the driver doesn’t take over the autonomous vehicle will try its best to navigate the situation. An example of level 4 autonomy includes Waymo by Google which has been navigating public roads since 2017.

Level 5 – At this level, the car will be completely autonomous and capable of self-driving in any situation and environment. Absolutely no human intervention is needed, to the extent where pedals and a steering won’t have to be a part of the construction of the vehicle.

We’re are still some way off from autonomous taxis where you simply tell the car where you want to go and it will take you, but several concepts are making the rounds from the likes of Volkswagen, Google and Audi.

SEE ALSO: "The foundations for automated driving have been laid"

Leaders in self-driving cars

For most automobile manufacturers, driverless vehicles will likely be at the core of their future strategies. Tech companies have also joined the race to be the first to get their driverless cars onto roads, indicative of the high stakes of this revolutionary market.

Here is rundown of four driverless vehicle projects to keep an eye one:

Waymo – The most widely known autonomous vehicle, the Google launched car has been in the game since 2009 and has racked up over 5 million road miles. Waymo has enjoyed an early lead in developing driverless cars and they aren’t showing signs of slowing down anytime soon and, is expected to pilot a fleet of driverless taxi-vans in 2018.

(Source: Shutterstock)

Waymo doesn’t manufacture its own cars and instead, it has partnered with Chrysler and Jaguar to provide the vehicles for its fleet. However, experts state that Waymo is far ahead of the pack in regards to technology.

General Motors – Similar to Waymo, GM plans to pilot a fleet of driverless taxis in 2019, having already submitted a petition to the US Department of Transportation to begin operating a commercial taxi service.

One big advantage that GM holds over Waymo is its ability to manufacture its own vehicles giving them the opportunity to perfect, which may prove instrumental in the long term. In addition, GM has received the confidence of investors, with SoftBank investing $2.5 billion in May 2018.

Baidu – Chinese search giant Baidu has 5 years of experience developing its self-driving software, and started testing its cars on public roads in late 2017. Baidu expect to have level 4 autonomous cars on roads by 2021.

Baidu is unique for its open source nature giving techies all over the world the opportunity to provide data and test the cars. The Chinese government has also given Baidu their seal of approval, allowing the company to test cars on public roads in Beijing, marking the first time autonomous cars will navigate the streets of China.

Volkswagen Group – Volkswagen Group holds the coveted title of having the only level 3 autonomous vehicle available for commercial audiences. The aforementioned Audi A8 can navigate at speeds up to 60km/h without the input of a driver.

The future of Volkswagen Group is looking equally impressive with signs that the company is looking to get the technology into the hands of the consumers as quickly as possible with recent partnership announcements of Nvidia and Uber. Volkswagen will develop self-driving cars with Uber which will utilize the powerful technology of Nvidia’s Drive IX SDK.

What are the benefits?

Safety – Road accidents are still a major leading cause of injury. An autonomous vehicle won’t suffer from bouts of road rage or the many other foibles that can hamper human drivers. Computers follow the rules and make the perfect drivers, and will likely result in a substantial decrease in road accidents.

A solution for disabled drivers – Automated vehicles will offer people with disabilities the ability to use a car, offering greater freedom and mobility, instead of having to rely on public transport.

Less congestion – Traffic that is dominated by autonomous vehicles will likely flow smoothly resulting in less congestion. Furthermore, there will be fewer cars that clog up streets and the spaces around building as an automated vehicle could feasibility drop you off at your destination and drive back to another location such as your house, to park.

Convenience – An autonomous vehicle will free a human passenger from the constraints of driving a vehicle, giving people the opportunity to work, read, sleep etc. while driving.

What are the challenges?

The obvious challenge to the autonomous vehicle industry is the potential dangers. We’ve already outlined the potential issues that may arise in regards to economics, but what about road safety?

Lost jobs – Every conversation relating to driverless cars ultimately begs the question, how many jobs will be lost? Theoretically, every job in the transportation industry is at risk, including taxis, buses and long-haul truckers. Driving instructors will also be a job at risk as the role could be replaced by AI instructors.

Want to hear more about AI and the risk of supplanting human roles? Check out the video below:
Cyber security - While physical road accidents can be significantly diminished in autonomous vehicle world, there is the issue of cyber security. What happens when an autonomous vehicle is hacked? Can an autonomous vehicle be tricked into seeing obstacles where there are none, for example?

Autonomous vehicles also operate processing large amounts of data which will be vulnerable to malicious use.

This is an issue that autonomous vehicle manufacturers will need to address. There aren’t any recorded incidents of a vehicle being hacked maliciously, but the first incidents will likely appear when autonomous vehicles hit the roads and become more common

Cost – Driverless cars are jam-packed with high-tech processors and sensors. As such, the price of a driverless car will be steep for most average earners, with costs expecting to exceed $100,000.

Loss of knowledge and an over-reliance on technology – If driverless cars become commonplace, people with the skills to operate a car will become less commonplace. This would be an issue in circumstances where the technology fails and a human has to take control of the vehicle.

Legal precedent – Autonomous vehicles are not 100 per cent safe. If an accident does occur, questions arise over who takes the responsibility. Does the owner of the car take responsibility, or does the manufacturer? Or even the developers of the artificial intelligence?
There are also questions to be asked about the regulation of self-driving cars and by what forces the future of the technology should be dictated by.

A better name for autonomous vehicles

What can we call cars that drive without human input?

At the moment, they go by a lot of different names, each one more verbose and clunkier than the last. We have; autonomous vehicle, driverless car, self-driving vehicle/car, robotic car, etc. Imagine telling a friend, "Don't worry about a lift, I have an autonomous vehicle coming to pick us up." You'd sound more than a little strange.

Ironically, the term 'automobile' would fit perfectly in this context, but that has already been taken. The day autonomous vehicles take over the roads, we will likely begin to hear a few different monikers and a natural vernacular form around the topic. But until that day happens, here are a couple suggestions for around the web.

Nemobile – It rolls off the tongue and makes sense linguistically. "Nemo" means nobody in Latin and 'mobile' is self-explanatory. You put the two together to get Nemobile. Short and sweet, this is one of the catchier ideas floating around, and you can also call them Nemos for short

Driver – Way into the future – if human drivers become a rare sight – the vocabulary centered around the act of driving will surely change. The word 'driver' may eventually refer to machines and artificial intelligence. Parallels can be drawn to the word 'computer'. Computer isn't a new term, its first recorded use was in the 17th century to describe a person who does calculations and mathematical equations. Nowadays, we only use the word computer to describe a machine that computes.

It also has a futuristic allure to boot. "I'll call a driver" or "My driver is filling up on gas" and "Oh no, my driver is stuck in traffic. Now I have to walk". Linguistically, it also makes more sense than 'driverless car' as every vehicle will technically have a driver, human or not.