Solving our traffic problems with self-driving cars?

Average reading time: 7 minutes

With the progress that has recently been made in computational power, connected technologies and artificial intelligence, the vision of self-driving cars is now finally coming true. Fully-autonomous cars are expected within the next two years. But the potential of this technology goes much further than just adding vehicle safety. It could just be the thing to solve many of our 21st century problems.


Much more potential than just cool looks.

While our cities get more and more congested and traffic gets out of control, we have to admit: our current model of transportation is not only inefficient but also dangerous. According to WHO, we suffer of 1.25 million traffic deaths each and every year. Why is that? And can self-driving cars be a solution? I want to argue that self-driving cars are not just a cool new technology – but the real completion of the concept of individual motorized mobility. In hindsight, the era of human-driven cars might be regarded as nothing more than a temporary transition era.

The mess of human-driven traffic

The abilities of an entity are usually defined by its purpose. Let’s regard the human body and mind for a moment from a strictly scientific purpose-oriented point of view. When we say, the process of evolution has shaped our body – with what purpose was it shaped? The survival in natural environment. Thus our primary abilities are based (not limited) on surviving. Our ability to react, our ability to estimate situations, our ability to sense dangers – all these are grounded in the two million years of human evolution. When we run on our feet and stumble upon an obstacle, we can react fast enough by mitigating the fall with our hands. We can sense the danger when leaning too far over a bridge’s balustrade. Our abilities have been formed to estimate heights, distances and dangers we encounter in our natural environment.

Cars on the other hand are a new concept. We have getting so used to shaping our world towards automobile mobility that we often forget, we are dealing with a technology that has only been around for a little more than mere hundred years. This is, needlessly to say, far too short for any human evolution to happen. We have to rely on our abilities, shaped to deal with far different challenges, to control them.


Today’s car traffic: A dangerous mess.

However, we do not have the abilities to measure the difference between 100km/h and 120km/h without even feeling any wind in our face. We can barely judge if 20 meters distance is enough safety distance at a certain speed or 30 would be better. We are tough challenged to observe the behavior of dozens of cars, bicycles, pedestrians around us while other thoughts might be on our mind. Can we brake fast enough in case something unexpected happens – and will the people driving behind us grasp the danger we are reacting to? When driving a car, we are pretending to have abilities which we simply cannot have.

Ultimately, a steering wheel is an interface. An interface between a machine – the car – and a control unit – us. This is the root of the problem: The control unit does not have the abilities needed to control a car in all possible circumstances – because the abilities have been shaped towards far other challenges in a completely different context. Wouldn’t it then be awesome, if we could prevent this incompatibility altogether and make the machine an integrated system – with a control unit that is actually made with the purpose of controlling a multi-ton super-fast vehicle?

The formal rules… and the informal

Self-driving cars are not a new concept. We have been using cars for 130 years. We have been using computers for many decades. Why did it take so long to build a self-driving car? Because it is vastly different from building any other machine.

Most machines we are building today operate in a controlled environment following pre-defined formal rules for their actions. The purpose of a self-driving algorithm however is to control a car in the real world. We sure have extensive regulations on how to behave in traffic, pretending to organize the relationships of all actors on the streets, but let’s be honest: Traffic regulations are only an illusion of control – and just a part of the abilities needed to drive a car.


Formal traffic rules are the defining parameters of the traffic system… are they?

Many modern cars today already master limited self-driving capabilities such as holding a lane. This is effectively a machine operating under a (fake) controlled environment. The driver tells the machine, when he’s on the highway, the traffic is easy and the streets and lanes are in good condition. However, for a car to operate really autonomously in the real world, the environment does not stop at the lane borders (if there’s even any). Unforeseen things can and will happen. New actors can anytime suddenly enter your environment. Other road users will improvise or simply make mistakes. Construction sites may spontaneously change the flow of traffic. There may even be places with no fixed regulation after all, such as the progressive shared space concept, increasingly being put to use in urban environments.

In the real world, traffic is governed by complex rules and relationships in which formal traffic rules only take a small place. In a traffic system dominated by human-controlled cars, a self-driving car being limited to just the knowledge of formal traffic regulations, is an alien.


Imagine a bus stopping in a crowded street. On the opposite sidewalk, a group of people are starting to run. A human driver may conclude, that the pedestrians most likely will try to cross the street in a non-cautious manner in order to reach the bus. As a result, he will lower his speed and increase his attention. This conclusion not only requires registering all potential actors around the car, but also interpreting their behavioral context.


Making context-sensitive predictions with limited knowledge.

The conclusion is that self-driving cars both need the ability to follow regulations and at the same time the ability to recognize human interactions, as fixed regulations are spontaneously altered. How can you even predict human behavior? Sometimes the conditions for certain actions may be different. Or the actions taken by actors may change. Example: As a traffic light is turning from green to yellow, some drivers will hit the breaks – while others will accelerate to pass before red light, sometimes well above the speed limit. To make matters worse, different actors are often influencing each other, eventually creating a complex system. Drivers surpassing the speed limit will lead to other drivers to act just the same. Drivers ignoring certain impractical rules will incentivize others to follow. A group of pedestrians crossing a street will incentivize others to follow, even if a car is approaching. So, how is a machine supposed to interpret human behavior, act upon dynamic rules that are being established on the fly?

Building a real self-driving car is only possible by the usage of artificial intelligence – a technique that has only risen to real success recently, and thus explaining why we could not build such cars in the past. A self-driving algorithm needs to learn from its experiences and incorporate a kind of adaptivity not possible for regular algorithms.

Training the A.I.

All approaches on true artificial intelligence incorporate a learning process. Well, what do we need to learn real-world traffic rules? Experiences in the real world.

Last month, Tesla announced, it had 780 million miles of semi-autonomous driving data, adding another million every 10 hours. Beginning in 2014, Tesla has equipped its cars with the hardware and software for its Autopilot. This is the incremental approach: Tesla is rolling out cutting-edge technology to its users and receives data when it is used. This way, the company has created an unmatched wealth of experience of its autopilot algorithm in the real world. Every experience helps in making the algorithm to get better. As updates are delivered via wireless connections to the cars, customers directly profit from advancements. Tesla’s CEO Elon Musk predicted the first fully self-driving cars to be ready in just about two years. On the other hand, of course, putting an in-development software on the streets, requires a high amount of responsibility of the driver, to judge what is possible with the system and what is not, and also a lot of self-discipline, to not get careless when being driven around (UPDATE in July: A Tesla driver died in an accident while autopilot was enabled. It seems to be both an error on the driver side, not paying attention anymore, and of the software, not recognizing a large white trailer on the highway against the bright sky. Tesla emphasizes that each driver has to acknowledge that Autopilot is an “assist feature” and does not provide full autonomy. For more info see Tesla Blog).


Tesla’s autopilot with the ability to change highway lanes. © Tesla Motors, Press Kit

Google uses a different approach, a more direct one. Their currently 58 self-driving cars – not even equipped with traditional steering wheels – drive through four selected U.S. cities using professional test “drivers” (or rather: observers), and gathered data of 1.6 million miles since 2009. While the Google approach obviously creates much fewer data than Tesla’s, it also allows for a more progressive and also more safe approach. When not having to worry about practical customer considerations, they can directly aim for full self-driving capability. As Google will need more and more experience data, the test program is expanded. Positions as a self-driving car test driver are currently open. (Update in July: Google’s cars now learned to predict cyclist behavior and hand signs using an A.I.-driven learning approach – very impressive! Source).

While both strategies have their advantages, it is foreseeable for both companies to have leading positions for self-driving algorithms in the foreseeable future. Traditional closed-environment small-scale testbed strategies can only advance to a certain point (mostly revolving around learning formal rules), but for fully autonomous driving and interactions with human-driven vehicles, extensive real world data is necessary. For example, Google is currently learning its cars when to honk – useful for situations of urgency and warning in the real world, not defined by formal traffic rules: “Our goal is to teach our cars to honk like a patient, seasoned driver. As we become more experienced honkers, we hope our cars will also be able to predict how other drivers respond to a beep in different situations.” (Source) Real-world data is key for learning processes revolving around self-driving cars.

The next step

In a world with ever expanding cities, how do we prevent the total traffic collapse? At a certain urban density, there is just no more room to increase the traffic capacity. The solution in fact could be self-driving cars. If we take traffic capacity as a limited good, traffic lights are always a very inefficient way of distributing this good. Intersections become bottlenecks. By using self-driving cars and thus removing the need for traffic lights, the capacity and efficiency of streets could dramatically be increased, possibly doubled, as Italian researchers have just recently shown.

Moreover, by combining ubiquitous connectedness with artificial intelligence, we could utilize intelligent traffic management to organize and spread daily commuting routes and thus make the most out of a cities’ existing infrastructure at peak times. If combined with the revolution towards electric vehicles, which is just getting started right now, and the expansion of regenerative energy sources, pollution could be drastically reduced and quality of life and health increased. Hopefully, self-driving cars are just the first promise to come true towards a more reasonable solution of mobility.

It is about time.


How can we improve mobility to ensure quality of life in the urban age?


Follow me on Twitter: @caladan42



Tesla becoming a mobility service provider. A whole industry’s future?

Average reading time: 5 minutes

Just recently, Tesla’s CEO Elon Musk released his second “Master Plan”, detailing the car manufacturer’s future development goals. And it did not fall short of expectations: Tesla in fact is not anymore just a car manufacturer, but it is about to become a mobility service provider – pursuing a development looking to transform the whole automobile industry.

Let’s remember: 10 years ago, Elon Musk published his first Master Plan. It laid out the grand strategy that we actually still see unfolding today: start by building expensive electric sport cars in low numbers for premium customers, use the margin to advance production and technology and eventually be able to mass-produce affordable electric cars: the upcoming Tesla Model 3. So in conclusion, the first Master Plan was basically a strategy of scaling – how to transform a promising start-up with a lot of ideas but much less abilities into a mass producer to deliver just the kind of product you aim for (this, by the way, is in some way very similar to Musk’s other big company – SpaceX).


The Big Picture

According to Musk (if we believe him) his efforts of electric cars are not just because they’re so popular and generate so much cash – they are not – but as means to a bigger picture: to contribute while humanity reaches a sustainable long-term solution for mobility, dealing with all of its current problems we observe every day. We have to realize that our model of mobility currently in use is not made to last and comes with great draw-backs: Pollution, accidents, inefficiency. To arrive at a sustainable mobility future, we will need to deal with all three of them.


The cliché of individual motorized mobility: Going where you want, when you want. But it comes at a cost.

Pollution is to be dealt with by electrification of cars, with the energy at best coming from renewable energy sources.

The accidents and resulting 1.25 million yearly deaths are to be dealt by increasing abilities of autonomous vehicles, surpassing the reaction ability of human drivers (A.I.-driven mobility might be the natural solution). Elon Musk wants to “Develop a self-driving capability that is 10X safer than manual via massive fleet learning”. Despite the latest setbacks, not only Musk, but everyone in the car industry expects fully autonomous vehicles to arrive in the next few years. BMW just announced its first fully-autonomous car for 2021.

But what about inefficiency? Now things are about to get really interesting. This might be the core message of Musk’s new Master Plan.


Taking the best out of individual and public mobility

Individual mobility is great. It offers the possibility of driving anywhere you want at any time of the day, every day – planned or spontaneous. The world transformation we have experienced in the last 130 years of individual motorized mobility is easily explained by looking at the freedom and utility privately owned vehicles offer to us. Cities today basically are formed by webs of streets – or in other words, cities are made out of pathways of individual mobility. The concept is that popular, we have shaped our world after it.

However the concept also has a huge drawback: It is totally inefficient. When we look at the streets, we mostly see cars with one or two persons in it, while the maximum capacity of five, seven or even more is almost never utilized. The results are not only pollution but also traffic jams, cost of infrastructure and generally decreasing quality of life both for drivers and citizens. Moreover, when we arrive at our target destination, the car just stands around, waiting for us to come back. Most cars are only used 10% of the day, being idle the rest of it and using up precious parking space in our ever-more congested cities.


Public transport: Efficient but unpractical.

On the other side, we have public transportation such as busses. They are great, because they move a large number of people at the same time, thus getting energy-efficient. At the same time, they offer a lot of mobility for people not owning a car. This is great because it does not fill up the streets even more and also provides some social equality – as in today’s economy, mobility is not luxury anymore, but a requirement. However, public transport naturally restricts the “when” and “where” and also has limits in utlity and comfort.

The core point of Elon Musk’s Master Plan therefore –this is my interpretation – is to combine the best out of individual and public mobility by adding fully-autonomous driving with managed car sharing to the equation. Linking multiple new innovations together leads to a solution that offers both freedom of travel and efficiency.


Progressing the Uber approach


Screenshot of Uber. (c) Uber, PressKit

It hasn’t been that long since Uber started a worldwide revolution of car sharing, allowing just about everyone to become a taxi driver. You press a button in the App and an Uber driver will arrive.

Tesla is taking this one step further. It will allow every Tesla owner to let his car drive around others while it is not needed by the owner. Or if you don’t own a car yourself, you just press a button in the Tesla App and a driverless taxi, owned by someone else or by Tesla itself, will arrive, taking you automatically to your destination. “Once it picks you up, you will be able to sleep, read or do anything else enroute to your destination.”, Musk says. Just like in public transport, you would not need to be attentive.

This concept would be very efficient, because it allows cars to be used all the time, and not just the 60 minutes of commute a day. If there would be an option to share the ride with others (on A.I.-optimized routes) – and I guess there will be – it would get even more efficient. But at the same time it offers a lot of the flexibility of traditional individual mobility.

From an economical point of view, the generated revenue will both incentivize the usage of the feature by the car owners and also create multiple additional streams of income for Tesla itself (via the platform itself, the appeal to users and the customers’ cost considerations) that traditional car manufacturers do not have access to. Car owners can let their car earn money for them while being at work or asleep.


The integrated strategy: Becoming mobility infrastructure

In the end, this means, Tesla is about to become a sort of infrastructure for mobility – with infrastructure certainly being the most influential position in an industry sector you can have (recently blogged about Google becoming digital infrastructure).

Beginning with this new Master Plan, and Tesla’s ongoing merge with SolarCity, Tesla is saying farewell to the time of being just an automaker. Instead, the company is aiming to become an integrated mobility service provider, spanning the full circle from energy generation (solar power on home roofs, SolarCity), energy conservation (batteries at home and in cars, produced together with Panasonic), energy consumption (electric-powered cars) to mobility distribution (new autonomous car sharing platform) and even recycling (recycled car batteries become home batteries for storing solar energy, Tesla Energy). The goal is clearly an “everything out of one hand” strategy. Building the actual car is only one part out of it. This is a revolution in an industry that has been focused on selling the best car product again and again for 130 years.


Cornerstones of Tesla’s long-term strategy. Mobility infrastructure instead of just a car producer.


A whole industry is about to transform

When we now look at the industry’s big picture, it becomes clear that the role of traditional automakers is challenged. And indeed, the revolutions of the likes of Uber, Google and Tesla have not gone unnoticed.

Toyota has just recently announced a strategic partnership with Uber. Volkswagen has invested 300 Million Dollar into Uber competitor Gett. General Motors has invested 500 Million Dollar into Uber competitor Lyft. Last year, BMW, Daimler and Audi together bought Google Maps competitor HERE for over 3 billion Dollar to be prepared for “future mobility”. BMW has also started an App-based car sharing service for certain models, including the electric i-series in major cities. Volkswagen is currently deciding on whether to build its own multi-billion $ battery factory, just like Tesla did.


These will become core sectors for automakers soon. A whole industry will look different.

These disruptions are tough calls for an industry petted in success for decades. It will be interesting to observe, to what extend well established automakers will be able to adapt to the challenges and how new players such as Alphabet’s Google, Apple or Faraday Future will perform.

But then again: Change in the mobility industry is more than welcome. We need to realize, that the current solution for mobility with all its downsides and inefficiencies can only be a temporary one. Even if we just want to maintain the quality of life we currently have, we need to find a durable, a sustainable solution. And by being sustainable, we can actually grow as a society.


The mobility of tomorrow will need to answer a lot of questions. But maybe we are just about to find that solution by linking several new innovations together.


Follow me on Twitter: @caladan42

Google’s Quiet Revolution: The double strategy to change the world.

Average reading time: 6 minutes

Google’s I/O 2016 conference has ended. What has the company announced? Some new Apps, a new Android version, some more virtual reality? Many people seem skeptical, some even are disappointed. Is Google step by step transforming from the Nerd Kingdom we know and love to a usual business company with the usual product cycles and slowly dropping in innovativeness – the seemingly final fate of each big corporation?

Not at all! I want to argue, that Google has a deep long-term strategy which revolves around two key points: 1) Becoming digital infrastructure and 2) Using this position to develope a true A.I.

Both points are interconnected. You cannot do (2) without (1). Let me explain, why I think Google and Alphabet are just getting started – and will be the spearhead of a development that is about to revolutionize our world. 

Google: The star is the software. And it is just getting started.



What differs Google and Apple the most – the two most valueable companies of our time?  I would argue it is around product placement. Apple does sell things, sell a brand. You go into an Apple Store and buy an iPhone, iPad, Apple Watch (or even an iPod). Apple’s most products are something you have to consciously decide to buy. You base your decision on the quality of the products, the image of the brand and the marketing. Things like these tend not to be stable at all times, hence the mild Apple crisis with dropping iPhone sales, we experience currently.

Google follows a completely different strategy. It does not want to be a product you have to decide to buy. It wants to be system immanent, the spider in the web. You don’t think about who owns the road when you drive to the next shop. You don’t think about who owns the power grid when you turn on your computer (well, most do not). This is infrastructure. Google wants to become digital infrastructure.


Ever thought about who owns this pole, before turning on your computer?

Google wants to offer services that people will use so frequently, without evaluating if one of the competitor’s service is any better. Or services, many people won’t even realize they are using. Google does not just want to be a brand. It wants to be the hub that sells and distributes other brands, and even a hub for knowledge, as Google’s many Open Source initiatives show. Google Search and YouTube are hubs for content and ads which other companies rely on to access the market. Android and its Play Store are the hubs for millions of Apps. Google Maps and StreetView are a hub for places, ratings and increasingly becoming nothing less as a way of perception of reality. Services like Gmail, Calendar, Google Photos and soon Google Home with its new Assistant are utilities designed to be used as naturally as our tooth brush. Why does Google all do this? Just to make money? Nope.


Artificial Intelligence (A.I.)

What is A.I.? Easy. When we use the term, we often mean cutting-edge technology. The “intelligent” car that automatically turns on the light when driving through a tunnel. Or maybe the “intelligent” computer player in a game? Without a doubt, this is not real intelligence, because intelligence is by definition not bound to limits and functions a programmer sets, but instead has the ability to learn from its environment and from itself – to become more than it was before.

A true A.I. builds upon neuronal networks. The same kind of networks our human brain is built of, with patterns and connections forming freely all the time. A true intelligence learns from the interactions with its environment – and thus is able to predict further interactions.

A real world example is the human socialization process. When we are born, we know nothing (much like John Snow), not who we are, who the others are, how to communicate with them appropriately and how their response will be. So in the process of growing up, we try ourselves out and by the feedback we get from our environment, we evolve and grow. Our random actions (output) leads to reactions (input) from the environment – be it positive or negative feedback – which leads to a learning process, and next time a further refined output.


Simple learning process: An intelligent individual is learning and refining its action through interaction with its environment.


Remember – we said, a real A.I. would be modelled after the human brain, a neuronal network. So how do you create an A.I.? You need it to evolve and learn. What do you need for a learning process? Input by the environment.

Well, how do you get input? By being the world’s leader as a digital communications hub, the infrastructure for communication and knowledge – Google. Google processes billions of transactions every day. Its search engine receives queries containing all aspects of human life and knowledge – from the birth of philosophy to the latest sports results and on how to have a good marriage, and can thus learn how all things relate to all other things. The same goes for all of Google’s services – whether being it photos, videos, voice queries and other user transactions. The A.I. learns through patterns of communication.

In the latest Google I/O conference, we learned that Google now has the ability to not only detect if a photo or video (without any written description) contains a dog in it, no, it can even determine which breed. How? Through A.I. learning!


Google’s position as digital infrastructure is perfect for growing an A.I. It can learn by billions of user transactions every day, and always gets better.


If you followed the I/O conference – or read any news site in the last days, you might have heard of the latest Google innovations. But the biggest announced innovation was a quiet one, presented at the beginning of the conference by Google’s CEO Sundar Pichai: The Google Assistant, a refined version of Google Now.

The idea of Google Assistant is basically to become your personal assistant, but including nothing less than all of Google’s functions and the combined knowledge of the mentioned billion queries each day. When you look at the upper blue model, being infrastructure is rather passive. You get used by others. But Google Assistant goes beyond that: Its active function is to predict what your interests and plans are – maybe even before you really do – and to support you to the best of its abilities. Want to go to the cinema? Google Assistant knows about your children and will suggest family-friendly films fitting to your schedule in your nearby cinema. Hungry on your way home, but in a traffic jam? Google Assisstant will suggest placing a food order with your favorite meal and time it to your estimated arrival. Couldn’t go to the stadium because of work? Google Assistant will let you know how your favorite team played once you are done.

This is much more like the original green learning model we discussed above, that every child has to go through when growing up (and every human continues to do so, albeit with lesser intensity). Google Assistant acts – it suggests things of its own and will learn depending on whether you agree on suggestions or not and what your reaction is. Its ability to know you and to predict your human behavior and needs will refine each time. This is true learning artificial intelligence.



Google Assistant acts actively and learns based upon the feedback it gets. Millions times a day.


When in March 2016 Google’s AlphaGo A.I. beat the Go world champion – something that was deemed impossible due to the complexity of the game (because you can’t just calculate what is the best move like in chess because there are so many possibilites)  – and even invented new strategies no human Go player ever had seen before, everyone recognized that Google had build a really good A.I. and deserved acknowledgement. But that was never the point. Go was never the point. It was a successful demonstration of neuronal network learning processes.

Now, what Google – and Alphabet as a whole – are doing, is using this small-scale tested A.I. learning technique and applying it on the wealth of information and communication the Google Universe is receiving every day. Apps like the Google Assisstant are not only an application for us to use and benefit – but also a measure for the A.I. to have better interactions with us and thus be better able to learn.

We live in the Age of Data, the Information Age. Due to digitalization, we are producing data with everything we are doing. Over 90% of all data that is stored anywhere has been produced in the last two years. But what are masses of data when it would take even millions of people all their life to read and understand them? Only a machine can understand this giant amount of data and act upon it. An intelligent machine.



Google is just getting started. With the strategy of becoming digital infrastructure – a hub for digital transactions and communication – and using this very data in letting a neuronal network-based A.I. learn, Google is the spearhead of a development that will revolutionize our world, in one way or another. While even today, the A.I. has more impact on our lives than most may realize, in just a few years, the A.I. will be a part of our daily life, like a toothbrush.

It is a quiet revolution.


So peaceful. Did anything happen?


Follow me on Twitter: @caladan42