The New Mobility World is where the future of mobility becomes reality. The big ideas, prototypes and projects of the first edition in 2014 have since become the status quo.
Ridepooling and ridesharing, connected and intelligent infrastructures, and first forays into autonomous driving services and buses. Although these ideas are currently available in many city streets they might be obsolete tomorrow. Today’s revolution is tomorrow’s convention. Mobility does not stop, it constantly drives innovation.
The New Mobility World unites these trends, developments and projects of the mobility of tomorrow under one roof. Experts, decision-makers and visionaries present and discuss topics such as artificial intelligence, solutions for climate change and air quality, smart cities or even infotainment systems and user interfaces. And so this year’s NMW in Frankfurt focuses on automation, connectivity, clean and sustainable mobility, urban mobility and mobility as a service (MaaS).
Over the next few weeks, we will present these key topics and provide a short introduction to each pillar of the NMW. Today we’re kicking it off with the automation of mobility: driverless from A to B.
Picture this: We’re on the highway. A car is driving at a consistent speed in the right lane. A truck appears on the horizon in front of the car. The car is faster and rapidly approaching the truck. As it gets closer and closer to it, changes lanes, pulls past the truck, and returns to the right-hand lane. What’s so special about an ordinary overtaking procedure on an ordinary highway? The car undertakes this manoeuvre independently; the driver dozes in her seat. The dream of autonomous driving has been with us for decades. But now we move from being a mere pop-culture phenomenon and science fiction vision to autonomous driving actually becoming a reality.
But not all autonomous vehicles are created equal. In fact, one differentiates between five categories:
The focus of the developers and manufacturers is currently on autonomous driving on the hybrid Level 4. It is simply more in line with the current state of technology. The reason: The tolerance for mistakes of fully autonomous vehicles in real traffic is extremely low. External influences such as the weather and complex situations such as an ambulance or the police chasing down the streets currently contain too many different variables and make complex demands of the systems. While Level 4 autonomy can already be tested and used on the road rather safely, Level 5 is still a case for closed ecosystems (test tracks, university campuses, etc.) where test conditions and external influences can be controlled. These problems are also taken up and analysed in an MIT study from 2018.
The rapid pace of development for vehicles with Level 4 autonomy is largely due to two different trends:
1: Machine Learning Algorithms and Computer Vision: The technical advances are the foundation for advanced AI, which can respond in a short time to various changes in variables in real life. The vehicles can correctly identify and solve problems or obstacles and learn from them.
2: The rise of ridehailing: The technology alone could not lead autonomous driving to its breakthrough. Only in combination with the rise of ridehailing could tech catch up to the market and vice versa. Why? Cost efficiency. Currently, the required components for autonomous vehicles – both software and hardware – are expensive. Due to the increasingly popular ridesharing services, providers only need to produce few vehicles for a larger amount of potential users. This way, the efficiency of the acquisition costs increases significantly. It is not without reason that Waymo plans its first autonomous vehicle fleet as a ridesharing service. Other ridehailing providers such as Uber or Lyft are among the leading developers of autonomous cars.
A prerequisite for the automation of all vehicles is the built-in sensor technology and the resulting data. The sensors are the eyes and ears of the vehicle and the data the neurons, which are processed as information by the software. Autonomous vehicles or assistance systems must constantly be aware of their surroundings, predict problems and react to obstacles. In the meantime, on average there are already more than 100 sensors in each mid-range vehicle. From 2021 on, the European Union will additionally only allow vehicles with automatic braking systems, lane keeping systems, intelligent speed adaptation, driver distraction monitoring, and event data recording.
Such developments are ausing the market volume to rise steadily. The market for LiDAR sensors alone – fundamental to autonomous driving – is set to rise from $230 million (2016) to $ 2.5 billion (2016), according to IHS Markit.
But not only sensors are fundamental to automated mobility. It also requires a form of artificial intelligence. Through digitisation, the car is evolving from a means of transportation into a computer on wheels. It must receive, process, understand and apply information within nanoseconds, which at the moment is only possible with artificial intelligence.
Machine learning describes algorithms that adapt and learn based on given data. Deep learning on the other hand describes digital neural networks that simulate scenarios, and interpret human decisions. Combined these to technologies produce said form of artificial intelligence. They make it possible for autonomous vehicles, intelligent infrastructure or assistance systems to be able to understand situations, react to them and learn from them.
The digitisation of mobility, as well as modern sensors, software, autonomous vehicles, and driver assistance systems, have major implications for road safety. With increasing prosperity in former and present “emerging economies”, the demand for mobility is growing there as well. The consequences: The global traffic volume is constantly growing and so are traffic fatalities. According to the United Nations, the number of road deaths increased by almost half between 2010 and 2020 to 1.9 million. At the same time, the fatalities in industrial nations have fallen rapidly over the past 30 years. One reason: digital driver assistance systems are becoming more extensive, more popular and a standard in new cars.
Nevertheless, according to the Bavarian Ministry of the Environment and Consumer Protection, 90% of accidents are caused by human error. In the opinion of Australian researchers, this figure is due to decline with the increasing adoption of driverless vehicle technologies.
Between 11 and 15 September, the centrepiece of the NMW will be the IAA Conference. Here the decision-makers, experts and visionaries of the mobility world have five stages at their disposal to discuss and engage what the future might hold. The call for papers is already on its way. Interested parties can apply as speakers until March 11, 2019:
Image Source: Waymo