If you believe that self-driving cars will one day solve your traffic problem, you have been heavily mislead
Three years lost. You are likely spend hours each day stuck in traffic, while you’d rather do so on friends and family. You also hate those crammed buses and trains you take when trying to ditch the car.
You’re not alone. On average, you will spend 80 minutes a day commuting to work (source), which makes it about three years of your daytime life, over a 40-year career. What a waste. Imagine you could re-employ half of it in personal time and half of it working: everyone’s total working time would increase about 8 per cent, and so would do each country’s GDP and we would all be happier and more free to follow our passions.
Google Car is not a meaningful solution:
If you believe that self-driving cars (SDCs) á-la-Google (let’s call them “G-SDCs”) will one day solve your problem, you have been heavily mislead. Tech giants such as Google, Tesla and Apple, but also traditional players like Volvo, GM and VW and many others are working on G-SDCs, as they all want keep or gain a place in the US$3.3 trillion global car industry. They make you dream of G-SDCs solving all your commuting troubles, ferrying you pleasantly, while you can enjoy reading, working, doing your make-up or clearing the first few emails of the day.
It is all a big lie. They are trying hard to get a foot in the door, but they will have a tough wake-up call when they will have to write off most of their investments, in about 10 years or less.
With a far more pragmatic approach, we could claim our three years of life extension much faster, in just a few years, rather than the decades required by Google’s approach. Here we will tell you how.
A driverless car may replace your Uber Black, with the same comfort
The real driverless world:
Assume for a moment that what they say is true: you want to enjoy benefits of SDC commute in total freedom, so you get yourself a G-SDC. Once you reach your workplace, what will you do of your car? Park it. That’s it? But the car can drive by itself, so it could do something far more useful than resting idle in a dark parking lot: it could pick up and drive other people. It will be particularly easy if those people can request the car through an app, as today is done for cabs.
Then why should you buy your own self-driving car, if you’d use it less than 5-10 per cent of its operating time? You would not. Then clearly G-SDCs will tend not to be owned by consumers but by a type of car-sharing company, and users will pay-per-use, as it happens for taxis in most cities.
One more step: what if instead of a car, there is a comfortable and spacious minivan, 10-seater, which along your daily commute picks up a few other people in your area and drops them to offices near yours, choosing a dynamically optimised combination of people and routes, extending your riding time by only 10-20 per cent, but allowing to reduce the price of your ride by 70-90 per cent. We bet majority of people would opt for that.
With scale economies, such a system should cost far less than a taxi, for two key reasons: 1) there’s no driver’s cost and 2) the cost of each ride is shared with several users at any point in time. At the same time, it keeps the comfort of a taxi, which picks you up at any point A and drops you at any point B, and it can be as spacious and comfortable as a high-end minivan. Furthermore, it will be always available, no matter what.
We once learned from taxi driver in Singapore that most of his colleagues will just stop driving on heavy rain – which is when you need them the most – as they are afraid of hefty down-payments they are requested in case of accidents. No, an autonomous vehicle would not refuse to drive. It’s autonomous, only in doing good things – for the moment.
“Wow! It seems amazing! A taxi always available at a fraction of the cost!” We bet most people would love such a thing. Let’s call this system “Closed Network – Personal Rapid Transit”, CN-PRT, or just PRT for briefness’ sake, to differentiate it from the G-SDCs.
The good news are not over. The PRT could be closer in time than you think. “When can we get it?”
Tomorrow. Or even today.
Your reactions: “Wow! It seems too good to be true! Where is the catch?”
Google Car is not a revolution
Most analysts say it will take another 10-20 years for mass adoption of G-SDCs. IHS forecasts only 230 thousand G-SDCs a year sold in 2025, a staggering 0.2 per cent of the total market. In 2035? 11.8 million. Another fantastic 9 per cent of total cars sold… in 20 years from now… Revolutionary? Not so much.
Your question: “Then how can you say that we will have this fantastic super-cheap and comfortable PRTs taxis tomorrow?”
Well, it’s very simple. The realistic and pragmatic self-driving vehicles of today do not need to be private cars at all and they do not need to be full SDCs or G-SDCs, how we called them. They just need to be public minivans. And they need a closed network of dedicated lanes in metropolitan cities.
Costly Artificial Intelligence:
There is one single daunting technology hurdle in G-SDCs today, which is the need for them to interact with the general human driven traffic. Humans are irrational most of the times, unpredictable, they have slow reaction times, they are un-disciplined, they cannot be programmed, they also use psychology while driving (how many times you avoided an accident, because you timely spotted that a driver was about to cut your way just from small nervous moves of the car?) that is very difficult for AI (Artificial Intelligence) to replicate.
Overcoming this hurdle takes decades of development and billions of dollars spent in costly AI research. If you dedicate one lane (yes, one single lane, that’s all it takes) in each metropolitan street for PRTs, you create a separate network of self-driving vehicles, which only interact with other similar vehicles, with fully programmed rules and no room for exceptions. The technology for such a system exists today, and it could be put in place already tomorrow. No need for complicated AI.
Then, why this seems not coming our way yet? Well, that’s not completely true, it is slowly being experimented, as we will see later, but it could come much faster.
Bold politicians wanted:
The reasons are multiple. First, no one has experimented such system on a large scale yet. Second, large investments would also be required to build such a system and to keep it running alongside the regular road network. It would be actually cheaper to ditch all current cars in a metropolitan area at once, and transform the general road network into a Closed Network PRT. In this case, in fact, only minor adjustments to the roads (e.g. light fencing, rationalisation of pedestrian crossings, installation of sensors and communication systems) would be required versus having two separate vehicle networks running alongside.
The network could be later on gradually optimised for the new system, even freeing up space for our outdoor life: think of current large avenues with four-five lanes, where only one-two lanes would be required for the self-driving vehicles, with optimised routes, no traffic lights and averagely seven-10 people in each vehicle during peak hours (vs. the current 1.55 of today’s cars).
But telling all car owners to ditch their cars overnight, and implement such a large system requires a move so bold that no politician would embrace right now, even knowing that Global Total Cost of Ownership (GTCO – including both private and public investment and running costs) of such system would be structurally and dramatically lower than current transport systems with much higher comfort and efficiency and lower environmental and social impact. This is pretty evident by itself, and maybe some university could do a comprehensive case study to show the enormous value of such business case.
It becomes even more evident when we touch upon social and environmental effects. In urban areas, vehicle traffic causes between 50 and 90 per cent of total air pollution, which in turn, causes 7 million premature deaths a year. Also, road accidents claim an additional 1.2 million lives and endless injuries with permanent disabilities every year.
So, should we forget about our dream city with taxis available at all time at the cost of a bus-ride? No. Not everything is lost.
It is coming despite everything:
This concept of public self-driving vehicles separated from the general traffic is also having its own experiments. The London Heathrow airport has a fully operational system with closed network autonomous pods taking you from the terminal to the car-park. In the Netherlands, the transport company Connexion has created its autonomous ParkShuttles and two campuses of the University in Wegeningen got connected with their autonomous WEPods.
In Singapore, last year, they had autonomous golf carts driving along pedestrian and cyclists in the large garden of Jurong lake for six days, which people could book through their smartphone apps. Singapore, in particular, aspires to lead the roadmap to the new paradigm of mobility based on autonomous vehicles, but it is not clear which view between the two different approaches the city-state is looking to embrace.
While the Minister of Transportation, in a recent speech, envisioned a system of autonomous vehicles in a closed underground network, at the same time it seems to embrace and support projects in G-SDC type of vehicles. We believe they are missing the “third way” we advocate as most immediately feasible, a surface closed-network PRT.
The more or less experimental PRT systems (actually the Heathrow pods are here to stay) are definitely less flashy, so they catch less media attention than G-SDCs, but unlike the latter ones they are real and working today, just waiting to be properly scaled up.
We expect more experiments along these lines will come from universities, parks, large factories and other large complexes. An even greater opportunity may come from China, where demographic expansion is so fast to induce real estate developers to build entire cities from scratch without selling a single apartment beforehand, making them apparently ghost cities, which developers expect to fill up all at once when demand converges. If you are building an entire city, spending several billion dollars in construction costs, you may as well spend a few more to make it the first “city from the future”, in an un-paralleled marketing gig. Sale prices would largely pay back the effort.
Tell your mayor:
It is clear that only a visionary and bold public administration would be able to take the plunge in designing the city transport of the future and realise it. Public opinion could help, in a “bottom-up” innovation push. So go and talk to your city mayor and your country government. Tell them this is the opportunity of their life to be remembered as the politicians who changed forever the face of our cities!
Leaders in PRTs, such as Ultra Global and the British universities which worked on the project at Heathrow, will have a dramatic advantage in tomorrow’s transport systems. They are not many at the moment, so this may be a more juicy sector for the giant of transport to invest, rather than in G-SDCs.
We are not the first one to think along this lines. Lindsey and Townsend shared a similar view on Quartz, highlighting how Google’s engineer are putting a lot of effort in solving the wrong problem, attributing this to short-sightedness. At the same time, we believe there is a different explanation to their apparent mistake: making self-driving pods in closed environments is relatively easy, thus would not give significant patent protection to a potential multi-billion dollar business, which instead a G-SDCs may give.
What is your view on the future of transport? Will we ride Google’s car? Will our cities re-shape in a giant Heathrow car-park? How long it will take, decades or just a few years? Share your view in the comments below!
Special thanks to Emilio Frazzoli and Paolo Santi for their comments on the early version of this article.
© 2016 All rights reserved http://ift.tt/1RGRXzw
The views expressed here are of the author’s, and e27 may not necessarily subscribe to them. e27invites members from Asia’s tech industry and startup community to share their honest opinions and expert knowledge with our readers. If you are interested in sharing your point of view, submit your article here.
The post You could live 3 years longer, and it is not for Google Car appeared first on e27.
from e27 http://ift.tt/1RGRXzy