Monday 29 April 2013

The Inevitable Rise of Autonomous Vehicle Fleets


Following significant further investigation of the potential impacts of the deployment of autonomous vehicles it appears that the overall impact on society could be even greater than that of the internet.  In a sense the internet was essentially a ‘virtual revolution’, whereas this will be a ‘real revolution’.  Why do I think this?

To understand this more clearly, then the paradigm shift itself needs to be understood and put in the context of the societal norms where it will be introduced.  For this post I will confine myself to the North American context.

Several of the special guests that have been permitted rides in the Google self-driving car in public roads have made similar comments along the lines of Coby Chase, TxDOT’sdirector of government and public affairs at the Texas Transportation Forum inFebruary 2013: “The remarkable thing was that it was a little unremarkable.”

Similarly the paradigm shift appears unremarkable – it is simply that a vehicle can travel unmanned.  But when we start to unravel this novel concept then we find that it has profound and remarkable implications for society:
  1. A vehicle that can drive unmanned can do work by carrying people and goods.
  2. A vehicle that can do work can make money for its owner.
  3. A vehicle that can make money will be in great demand in a free-market economy.
But again, this initial unravelling doesn’t reveal the breadth of the impact of the autonomous vehicle.  So allow me to sketch out a possible implementation scenario that will unravel the new paradigm a little more.

Once the autonomous vehicle is certified safe for unmanned use then a number of businesses/sectors will be more than ready to purchase this technology.  In fact I expect that their orders will have been placed many months or even years in advance, as if they don’t utilize the cost saving and efficiency benefits of the autonomous technology then they will lose out to their competitors:
  • The trucking industry – by removing driver costs, reducing fuel costs and reducing maintenance costs then they can maintain profit margins and still the price of goods in shops will reduce.
  • The taxi industry – any good taxi driver knows this technology and ‘this day’ is coming.  For a New York Taxi the driver is approximately 57% of the cost of a ride – it is difficult to see how they could compete with autonomous taxis.
  • Car rental companies – most users will quickly realize that that hiring an autonomous car only when transportation is needed will be cheaper than a longer hire of an ordinary vehicle that is likely to be unused most of the time.  Also their rental autonomous vehicles will suffer less damage, require less maintenance and overall be cheaper to run and allow them more flexibility on not requiring returns to a specific location.
  • Car-share companies – their business models will naturally migrate to autonomous technology as it a simple progression of their existing models.
  • Ride-share companies - their business models will naturally migrate to autonomous technology as it a simple progression of their existing models.

You will note that a common theme is emerging here.  The early adopters all run fleets.  But what will happen is that (apart from the trucking industry) their business models are converging – they will now in fact be competing against each other.

But what about the average person?  The more entrepreneurial minded individuals will realize that with an autonomous vehicle that they will be able to use it for their commute to work, but they can then assign it to an ad-hoc autonomous taxi company operating in the ‘cloud’ that will, for a small fee, hire it out to those in need of transportation, at cheaper rates than even the taxi companies.  Users in need of a vehicle simply use a mobile device such as a smart phone or Google Glass etc. and send the details of their travel requirements to the cloud. Before the time that the vehicle is due to be returned to the owner then the cloud based company can have arranged for any maintenance, cleaning and re-fueling.  Thus the owner might actually make a small daily profit, even allowing for depreciation – which is a considerably better financial situation than an ordinary car which sits idle for around approximately 95% of the day.

Again – the fleet theme is repeated.  This time with private individuals and their low overheads and low profit expectations against the aforementioned business based autonomous vehicle fleets.

But what about public transport?  Well it is easy to see that bus services could be severely impacted.  They require riders to travel to and from fixed bus stops and will therefore have a lower level of service and probably have a lower quality feel than using an autonomous taxi.  Those buses serving high density corridors will always have the advantage of being able to densely pack passengers into a single ‘metal box’, but on any routes where road space is not at a premium then some, if not all ridership, could be lost to the autonomous taxis.  This leaves the bus operators with an interesting dilemma – of how to adapt their operational and business models to survive, or even thrive in this new environment.  The use of autonomous buses is certainly an option, but research is definitely needed to determine what might be an optimum solution.

And LRT?  Again the principle of high density corridors ensures the continuing need for LRT, but the lower-ridership peripheral routes may need review as to their continued viability.  What is of concern to the fiscally minded, is whether the operational, business and revenue models for proposed LRT lines or extensions are sufficiently robust for their plans and designs to continue being designed from within the existing paradigm.  When the large capital costs of LRT construction is taken into account, and the operational subsidy that most service require, an autonomous taxi alternative, funded by the private sector, may begin to look a very attractive alternative.

So there appears to be market forces at work, because autonomous vehicles can make their owners money, that could lead to rapid deployment and a certain degree of market penetration.  But there is another very significant market dynamic, or trend that will come into play as well.  That is the rapidly growing trend of the ‘shared economy’ which is well illustrated by the rapid growth of carshare and ride-share services, especially in trend-setting hubs such as San Francisco and other Californian cities.  This is clearly seen in the statistically significant reduction in ownership of cars by the younger demographics and the rising average age of gaining a driving license and the fact that they simply don't drive as much as in the past.  Much of this is related to greater awareness of environmental and sustainability issues from education, as well as a growing addiction to social media – where driving has now become the distraction.

Proof can be found in the claims from carshare companies, such as Car2Go who operate in Calgary who in conversation with myself claimed that a single Car2Go vehicle can replace up to twenty privately owned vehicles.  A review of research shows that one car-share vehicle can be seen to remove between nine and thirteen other vehicles from the roads.  This principle can be expected in autonomous vehicle fleets, and a study by the Earth Institute (EI) of Columbia University “Transforming Personal Mobility” indicates that in a successful autonomous fleet one autonomous vehicle could replace approximately six private cars.  In addition, the EI authors found that by relinquishing private car ownership that the average person could reduce their annual transportation costs by approximately 40% when using conventional cars as the base for the autonomous vehicle.  But when ultra-lightweight electric powered autonomous vehicles are used then the cost can reduce by up to 80%.  For an average person these savings could be significant multiples of their current disposable income and could result in substantial quality of life improvements.

So overall it appears that there are safety, efficiency, financial and environmental benefits for users to switch from privately owned cars to autonomous vehicles, but even greater advantages if they relinquish ownership of private vehicles and use fleet autonomous 'taxis'.  Those that continue to drive themselves will actually be sub-optimal road users in a number of situations, especially in dense traffic as for safety they should maintain larger headways due to slower reaction times.

How could this all affect our cities?  Well this is where we see a very interesting challenge emerging.  Most North American cities it seems would like to make their streets much more liveable and desirable places to be – hence the rise and rise of concepts such as ‘reclaim the streets’ and ‘complete streets’.  The desire is for pedestrians and cyclists to be actively encouraged and to remove as much fast-moving, dangerous and polluting traffic from urban streets as possible.  We see so many opportunities to move towards these ideals with the deployment of autonomous vehicles.

Firstly the requirements for parking will reduce dramatically as the autonomous vehicles can simply drop riders off and then either proceed to free parking outside of the inner city area, or be available for the next hire through the cloud.  This gives rise to the interesting question of ‘What do we want to do with this reclaimed land and these re-claimed parking structures?’.  We suspect that the urban planners and the private developers could have diametrically opposite desires here – which is why it could be very important for city planners to review policy at an early stage.

Secondly, because of the efficiency that autonomous vehicles will move through inner-city streets, as mentioned earlier, the human drivers will be highlighted as the sub-optimal element.

Thirdly, because autonomous vehicles will be the most courteous and safest of drivers, the opportunity to promote pedestrian and cyclists facilities above autonomous vehicles will be an enticing possibility for urban planners.

Finally, as autonomous vehicles won’t crash as much there will be a desire for them to shed up to three quarters of  their weight.  That is the weight that current vehicles carry simply because we require that they protect us in the case of a crash, which 95% of the time will be as result of human error.

When these factors are combined it is possible to identify that there would be a growing desire to ban human drivers from a city core and create something like the London CongestionCharge Zone, where only autonomous vehicles are allowed inside the defined zone.  With the ideal conditions to optimize autonomous vehicle fleets we expect that ultra-lightweight electric autonomous vehicles could become the standard vehicle to journey within the autonomous vehicle only zone providing safety, operational efficiency, financial and environmental benefits.  Pedestrians and cyclists would feel much more secure than with human drivers and the possibilities to improve the streetscape and promote community living and improve quality of life could have urban planners in some form of 'planning heaven'.

If Google do release their autonomous vehicle technology to the public in 2018, and autonomous vehicles are certified safe for unmanned use in say 2020, then taking a very optimistic view with this technology I predict that the first city might institute an autonomous vehicle only zone possibly as soon as 2023.  We may even see a race for the first city in each country to implement such a zone as the benefits could be very appealing to both city centre businesses and residents.


Tuesday 16 April 2013

WOW! - 96,000 Miles Without Safety Critical Intervention

I have just seen the ITIF Talk 10 April - At around 29 mins of the video, Chris Urmson of Google confirms that they have done in excess of 500,000 miles testing of the Google Self-Driving Car on the public roads, but the headline for me is that they have done 96,000 miles without safety critical human intervention.

Let's put that in perspective:
Extending Bryant Walker Smith's calculation from here:

The average person has a crash about once every 110,000 miles.
Previously Google told us 50,000 miles without intervention, which meant there was only a 27% probability that the car was as good at avoiding crashes as person.
But now, with 96,000 miles there is a 45% probability that the car is as good at avoiding crashes as a person. That is maybe 6 to 10 years for an average driver without a crash.....

Quite simply - Statistically the Google SDC in the 'ideal' driving conditions of the southern States is almost as good as the average person.

But I assume that we need at least 95% confidence that they will crash less than people - which we won't reach until 473,000 miles.

On DriverlessCarHQ we counted 32 unique license plates for Google SDCs - and we estimate that Google could possible be racking up somewhere around 1,000 miles/day of testing on public roads. We have no idea how much simulator testing they have achieved - but clearly any improvements seem to be feeding back well into the real world testing.

Whatever the metric that NHTSA require to certify the SDC safe for public use in the U.S., it is clear that Google are making very solid progress in their development program.

WOW! - Google are indicating that this technology will be in public hands in about 5 years time - so from when they first said this in Sept 2012 we can maybe expect 2017-2018.  That sounds very plausible given that their development program only started in 2009 and four years later they have a self-driving car that is already almost as good as a person statistically speaking.  

With another four or five years of development and the potential to rack up probably another 2 million miles of testing then this whole project looks very credible and on program based on my own interpretation of the data available.

As ever - what do you think?  Please add a comment and let me know.

Sunday 14 April 2013

Urbanism Speakeasy - Driverless cars, the most courteous drivers

A few weeks ago, one of my Institute of Transportion Engineers contacts, Andy Boenau invited me to do a podcast for his Urbanism Speakeasy website.

Andy is doing a whole series of interviews on all sorts of subjects around transportation aimed at listeners from all backgrounds, with a view to avoiding jargon and overly technical explanations - to make our fascinating transportation profession accessible to all.

Please let me know if we succeeded with my interview - I should warn you that it is 45 minutes long - so maybe download and play in a quieter time of day.  There is plenty to think about as we covered a whole range of subjects in the the autonomous vehicle space - which will affect almost every aspect of society

You will find the podcast here.





Thursday 11 April 2013

An elephant in the room?...


On Wednesday 10 April, 2013 I had the privilege of presenting at the Canadian Institute of Transportation Engineering (CITE) Annual Meeting in Calgary.

I attended some really excellent presentations on the Monday and Tuesday.  Many of the projects or designs being discussed involved expensive infrastructure and long range plans looking forward 20, 30, 40 or more years.

But in almost every presentation I had the frustrating thought: 'This project or design would have looked different if the engineers and planners had been aware of the impact of autonomous vehicles'.

This resonates very much with the Freakonomics challenge of December 2012 that had spurred me on to start my ITE blog:

"Surprisingly few people, 
even within the transportation planning world, 
are talking about this pending revolution." Freakonomics

So when it was my turn to present, I started with a slide asking "Is there an elephant in the room?" and explained that I would like to move the discussion forward so that ITE members can overcome the credibility gap and progress along this scale of being:

  1. Unaware
  2. Informed
  3. Challenged
  4. Excited
  5. Active
In the rest of my presentation, which can be found here, I explained:
  • how technology is developing exponentially
  • how autonomous vehicles work
  • the level of development of autonomous vehicles
  • details of the Google Self Driving Car project in particular
  • the paradigm shift that occurs once autonomous vehicles are certified safe to travel unmanned
  • benefits and challenges
  • when we can expect this technology to arrive - 2017-2018 according to Google
  • the transformative effect it could have on transportation and society
  • a possible implementation scenario
  • possible public and private sector responses
  • challenges to transportation professionals and our projects
During the 90 minute round-table discussion that followed the three technology themed presentations we talked about many transportation issues that arise from the deployment of autonomous vehicles.  But I was not surprised to find that the discussion frequently strayed to touch on many topics outside of transportation, as those present gained a greater understanding of the magnitude of change that autonomous vehicles will bring.

I talked about unintended consequences of not planning for autonomous vehicles, as I was not aware of any jurisdiction in North America that has currently allowed for the impacts of autonomous vehicles in any of its short, medium or long range transportation plans.

But one unintended consequence was that several conference delegates came to the mic to contribute or ask a question, and basically said 'Paul, you've scared us!' - whereas I had hoped to merely 'challenge'.

Perhaps there wasn't an elephant in the room after all.  Maybe it was an 800 lb gorilla!