Autonomous Cars – An Overview

Towards The Edge

Autonomous Cars are a big thing.

They are the best example of a revolution, which is hitting us at full speed: The new decentralization wave, also known as Edge Computing.

Nowadays automobile manufacturers are ridiculed as old economy by the tech scene (a.k.a the new economy).

However, the ongoing revolution clearly plays into the hands of the old economy.

A shift of power is imminent.

Data is the oil of the 21st century

For the reason that data is being created directly in the products of the automobile and engineering industries, the cards are shuffled anew.

After all, the keys are now in the hands of the Old Economy. One autonomous car will create 40,000 Petabyte a day. The front video camera alone generates 300 gb/h.

Most noteworthy, in the digital age, the one with data earns the money.

Not only Tech Behemoth Alphabet recognizes this shift towards the edge.

Countless tech companies are pushing into the field of self-driving cars – with big investments. China is giving full throttle.

This article will name the most important components of self-driving cars. In addition, it distinguishes between the expansion stages and shows where we stand today.

Components of Autonomous Cars

First, an autonomous car needs many additional sensors.

Waymo Chrysler Driving

1. Video Camera(s) – Surveillance of the environment and detection (lights, traffic signs, etc.)
2. Radar Sensors – Monitoring of environment in front of the car
3. Lidar – Monitoring of environment 360° via light and ranging
4. GPS – Positioning
5. Ultrasonic Sensors – Detection / Measurement of close objects
6. Odometry Sensors – Positioning

So, if you wondered, where all the data comes from – now you know.

Next, we will have a look at the different levels of autonomy.

Five Expansion Stages

Besides Level 0, which means no automation, there are five different levels:

• Level 1 – Driver Assistance
  Today we see many cars, which are Level 1 already.
  Level 1 cars need at least one driver assistance feature like adaptive cruise control or lane-keep technology.
• Level 2 – Partial Automation
  To be classified as Level 2, a car needs at least two advanced driver assistance systems (ADAS).
  Extremely important is a sophisticated coordination between the features like automatic emergency braking.
  The driver is still responsible. Examples today are e.g. the Tesla Autopilot or Mercedes Distronic Plus.
• Level 3 – Conditional Automation
  Now things are getting interesting.
  A car on this level must be able to drive autonomously – under good conditions.
  Good conditions are what the car is designed for, e.g. driving in a city with a maximum of 30mph.
  Yet the driver is still responsible.
  Google’s test vehicle was able to achieve Level 3 in 2012.
  Audi is currently trying to obtain legal approval for its conditionally automated Audi A8.
• Level 4 (High Automation) and Level 5 (Full Automation)
  First the bad news: There are no Level four or Level five cars available, as I’m writing this article.
  A highly automated car (Level 4) is able to drive without a driver.
  Nevertheless, it underlies certain restrictions like speed limits or geofencing. A driver can still take over.
  Level five cars will have no human controls inside.
  These cars are able to do whatever humans can do with cars today. No Limits.
  Experts judge that we are decades away from a fully automated car.

New Opportunities

In summary, it is clear that modern cars are fully equipped with high-tech. However, there is still a long way to go towards self-driving cars with a high or full automation level.

Nevertheless, we are experiencing a big change. More and more data is going to the edge.

Moreover, so are the new business models.

It will be very interesting to see who will be successful on this stage.

All the big players around the world like Alphabet, Alibaba, Volkswagen and Toyota know the potential.

We will see how the new kings of the edge will emerge. The kings of the Internet could disappear into insignificance.

Moreover, this article only covered one example – the car.

With the Fourth Industrial Revolution picking up speed countless more use cases appear.

Just think about all the connected machines and sensors in the various industries.

Exciting times lie ahead.

Header Image Source: pexels, pexels license
Article Image Source: wikimedia, CC BY-SA 4.0

About Post Author

Fabian Schmidt

Fabian Schmidt is working as Head of Software Engineering in Manufacturing at SICK AG. He has an extensive IT background and has a faible for lean and agile organizations.

See author's posts