Automotive Business Intelligence Trends & Vision

A Closer Look At The Current State Of Autonomous Driving

Cars Tokio Highway

Self-driving cars have been a vision and dream for decades and at some point over the last years we all recognised that we have reached a point of technological advancement that makes autonomous driving possible. However there is a huge BUT. Not only is it one of the most complex tasks to integrate sensor-based and highly rational technology into the emotional and imperfect traffic world of human drivers it’s also a heavy challenge to find solutions for legal, social, economical and ergonomic conflicts.

In order to get a clear picture of how close to a breakthrough solution we actually are as well as to understand the topic we need to take a look at autonomous driving from various angles.

According to a recent  J.D. Power Vehicle Dependability Study the number of problems with infotainment, navigation and in-vehicle communication systems (collectively known as audio, communication, entertainment and navigation or ACEN) has increased and now accounts for 20% of all customer-reported problems. ACEN is now the most problematic area on most vehicles and is the cause of the industry’s 3% year-over-year decline in vehicle dependability.

“The increase in technology-related problems has two sources,” Renee Stephens, vice president of U.S. automotive at J.D. Power, noted. “Usability problems that customers reported during their first 90 days of ownership are still bothering them three years later in ever-higher numbers. At the same time, the penetration of these features has increased year over year.”

Fact is that humans are very different and that if we develop new electronic features for the driver there is a good chance that we will create a solution which won’t fit everyone. One of the best examples is the BMW iDrive controller. When it was introduced in its first version a couple years a ago there have been very different reactions. From “useless joystick for kids” to “innovative and ambitious concept” we read everything in the automotive press. A couple years later and after several steps back with reintroducing several buttons in addition to the controller the iDrive is now an often copied and successful infotainment interface.

How does this correlate with self-driving technology? Drivers need to trust a system, this is even more true for systems which not only control the infotainment but driving behaviour of the car. The results of the J.D. Power study clearly show that brands and manufacturers need to build up a lot of trust before drivers will take their hands off the steering wheel.

A possible strategy to handle this challenge could be the introduction of autonomous driving features in small pieces. Again, BMW offers a great example with the new 7 series model which can be equipped with a variety of “advanced driver assistance systems and automated driving functions”. Among them are partially autonomous extras including a program which allows the car to enter and park in a garage on its own.

BMW 7 series
The new BMW 7 series enters and exits a garage without driver (image: BMW)

Looking at driverless cars from a social perspective we have to ask ourselves: If we don’t drive cars anymore, then why do we need them? Traffic jams, pollution, noise and high human activity – in big cities cars are a concern in many ways. Due to limited space streets would be the perfect, existing infrastructure for new public transportation concepts based on autonomous driving technology. From low-cost self-driving buses to premium cabs – there are a lot of scenarios in which shared transportation could make sense.

In rural areas on the other hand, driving is often enjoyable which raises the question: why would I want to give up driving myself?

The above shows that there are different markets and needs and that the introduction of personalised, self-driving products could fail due to better living in cities and the missing need for supported driving outside of cities.

Licensing a self-driving car in the US is comparably easy however additions to the legal framework in California show that the high number of autonomous vehicles on the roads of the golden state have become a concern. New requirements for “Public Deployment of Autonomous Vehicles” therefore include the need for a steering wheel and a driver who can take control of the vehicle at any time.

In European countries laws and guidelines are introduced just now but are also very conservative. In Germany, the German Federal Minister of Transport and Digital Infrastructure, Mr Alexander Dobrindt presented in September 2015 Germany’s “Strategy for automated and connected driving”. The comprehensive document including law, infrastructure, innovation and data protection seeks to reach the objective of making Germany a leader in automated transport and connected vehicles. However the document shows that autonomous driving will require way more than only sensor loaded cars – e.g. it talks about high speed mobile internet coverage as well as special driver training and cyber-security.

Besides those challenging circumstances self-driving technology is also not yet ready for our imperfect world. In a study, Google, which logged 424,331 “self-driving” miles over the 15-month reporting period, said a human driver had to take over 341 times, an average of 22.7 times a month with bikes, pedestrians, other cars and tree branches among the scenarios the robot cars couldn’t handle. Lately one of the Google cars even caused a little accident when it underestimated the behaviour of an approaching bus.

Google Car
One of Google’s self-driving cars – the cute vehicles are still learning to deal with human behaviour (image: Google)

Google now will re-programm the cars in order to make them understand that big vehicles are less likely to break for them. However, there are tons of other scenarios which humans can handle due to judgement which are not yet part of the self-driving vehicles’ program.

If we further imagine a mixture of autonomous and common cars on our roads it might become hard for humans to judge as driverless and human-driven cars might react in different ways.

To put all of this in a nutshell, self-driving cars are still in a research state; however we are already in the stage of field tests. It will take at least a couple more years until we will see a completely safe and reliable system as well as to establish a roadmap on how to integrate and use autonomous driving technology in personalised driving and public transportation.

At the same time advanced driver assistance systems and automated driving functions will continue to move in to our cars in order to make driving more secure and comfortable.

To excel in the autonomous market we join and support working groups and R&D projects, quite recently we started to support a R&D project for next generation environment sensing systems, with artificial intelligence (AI) and functional safety features (ASIL C/D).  For more information or support with automotive technology and systems feel free to contact EBV here and to learn more on our dedicated Automotive Website here.