Part 11 - China, virtually in the drivers seat

In the last few blogs I have  been looking at how things are shaping up in the world for the implementation of this exciting new technology.

What has become obvious is a lack of coordination and cooperation. There doesn't seem to be a concerted effort to make this happen. Governments are doing things on a state by state basis and while a lot of talk is going around about national standards in many countries this does seem to be a long way off . Manufacturers are working on systems independently of each other, as would be expected in the early days of development but standardisation of technology allows economy of scale in manufacturing - required for a wholesale roll out of this across entire countries.
There are many other issues as I discussed last week holding back the technology - at least in the west.

I have written a few times about what is happening in China, and the more I look the more I see things accelerating.

One of the key things in making driver less cars available is the human attitude. In China a recent survey found 75% of people would like to drive in an autonomous car but in the US it's a different story with 75% of people fearing autonomous cars.

China has a lot of reasons to want this technology to work and is using the opportunity to couple this with the development of electric vehicles.

Given that china has massive pollution and congestion problems and estimated road deaths of 200,000 a year, they have a very strong incentive to make this work, and more importantly, because of the political structure they have the ability to make this happen.
More importantly, from a world perspective, they have the ability to make this happen very quickly.

The west is agonising over testing and many experts say that the testing phase will last for many years and untold millions of kilometres before governments allow it legally on the road.

However there is nothing like the real world to put things in perspective.
This is illustrated by an interesting article from the New Yorker which talks about Tesla gathering masses of data from their users driving on autopilot.
This data collected from Tesla in a single day is equivalent to the total mileage gathered by Google cars over the entire testing life - every single day. 
Just to be clear, Tesla cars are logging over  million miles a day in autonomous mode and that is providing free real world data back to Tesla.

Now, imagine the data that can be collected from fleets of cars on the road in China.

Is China leading he world with self driving cars?

China has already come up with a draft road map for the next 3-5 years to have self driving cars on the road and Li Keqiang, automotive Engineering Professor at Tsinghua University is leading the committee.

The draft will define standards including a common language for vehicle to vehicle and infrastructure to vehicle communication. 

This committee has the backing of the powerful government Ministry of Industry and Information Technology which give legitimacy to the proposed standards and a fast track to making it happen.

China also has a raft of new car makers springing up including Future Mobility Corporation Limited.
For a new company it has made large first steps by hiring top BMW designers to work on their cars.

There is a push happening and this article by the Chairman of Zhejiang Geeley shows the reasoning behind the push and how seriously it is being taken. 
From a financial perspective it makes great sense as well and the Boston Consulting group says "China will be the largest self driving market by 2035 taking up to 30% of sales that year."

China leading the push - is that a bad thing for the rest of the world?
Yes and (maybe) No.

If China has a large scale roll out of this technology in the next few years, which is entirely likely, then the standards implemented, refinements made from lessons learnt and the sheer prestige of having these fleets in operation gives them a huge edge in marketing to the rest of the world.
Not to mention the amount saved by using home grown technology rather than importing it.
Good news for China, not so much for the likes of Ford and GM who are struggling to get their test cars on the road due to the lack of standards, laws and legal liability.

On the flip side, there will be a huge amount of technology and data coming out of China which could be utilised to the wests advantage to leverage their own testing and development.
This is if the boom doesn't create a brain drain to Asia and leave the west trying to play catch up.

What this could be is a major shift in technology from the west to the east which from our perspective may not be entirely desirable.

Other Asian countries aren't sitting still either with Singapore to test its first fully automated taxi this year on the streets and wants a fleet running next year.

This is a simple statement but think of the ramifications.
A city state running automated taxis within a year! Johny cab is here....

In a state where owning a car starts at more than $200,000, then a car as a service makes a whole lot of sense. When you consider that they could have 300,000 auto cabs on the road to cater for the population and not wait more than 15 minutes for a ride then it does indeed make a lot of sense.

Last week I ended with the statement Automated vehicles - Who's driving?

Now it may make more sense to say "China - virtually in the drivers seat"

It will be very interesting to see what comes up at the Beijing motor show that opened this week (2016).

Part 10 - What could possibly go wrong?

As anyone who has read any of my previous blogs will recognise, I am quite passionate about autonomous cars.

For the sake of argument this time I am going to play the devils advocate and see what will be the obstacles and what could slow the implementation of this technology.

Firstly a couple of notes.
As most people these days accept, autonomous vehicles will happen.
China for example has massive transport issues and sees this as a way out, probably also giving them more control over the populace at the same time but that is another issue. They will pursue this much faster than the west will because they can do so far easier and the west will be watching and learning from their problems.
More than 90% of vehicle incidents causing death or injury are caused by the driver.

Ok so what could go wrong?

The major issue will be people. Many people inherently distrust technology and a very real human trait is to resist change. Fear of the unknown means that education has to be the key if government wants to implement this technology and by building up the main advantages (saving lives) and guaranteeing the reliability of the software would go a long way to addressing this. People need to feel comfortable that they can watch a movie, read a book or work in the car, secure in the knowledge that they are not going to end up as an impromptu barbecue on the side of the road.

Once this tech hits the road we have to  do everything in our power to ensure that the technology is so good that the road toll reduces significantly before there is a death that can be attributed to an automated vehicle because if that happens early it will be used to discredit the whole concept.
Undoubtedly deaths will occur, no system is perfect, and of course no road death is acceptable.  When the road toll is reduced by a very significant amount we can learn from those unforeseen incidents that do occur. The great advantage is that we will be able to do something about it easier than we can now as the data from incidents can be analysed in depth. Each and every death or injury will mean that others won't suffer the same fate as we find out why and stop it happening again.

Liability. Some governments are tackling this now, the UK is putting frameworks in place and talking options on insurance and liability already so they will have that in place before there is an explosion in affordable technology in this field over the next few years and people startg to seriously use it.

The US appears still to be on a state by state basis with conflicting laws being formulated. Obamas' new 4 billion dollar initiative hopefully will coordinate state and federal efforts to accelerate the technology. If successful there will be a federal overview which could allow national standards and all states and the federal government hopefully would be on the same page.

Here in Australia there appear to be very little happening at all. South Australia seems to be the only state looking at this seriously and is changing the law to allow testing on the road and are discussing liability. I don't want to keep flogging a dead horse but Australia really needs to have a coordinated approach or we could seriously lag behind the rest of the world - and South Australia...(WA humour).

Ethics, I personally don't see this as an issue due to the nature of the technology but there are many articles that discuss this in depth. The theory is interesting but if viewed in a real world and compared to the current reactions of drivers in dangerous split second decision scenarios, I for one, would be far happier for a vehicle to react for me. Given my reaction times and speed of thought, everyone would have a better chance of survival.

Cost. The cost of the technology may be a deciding factor in the initial uptake of the technology but it will drop as volume increases and it becomes more widespread as economy of scale kicks in. One of the major contributors to that will be the "Taxi" industry, if you can call it that, where Uber, Lyft and a raft of others will be early adopters and pour billions into having this available so they can get a lead on everyone else.
What they would like to see, and I believe is a reasonable assumption, that private ownership of vehicles will decrease as the "car as a service" becomes very affordable.

Government action. I feel this to be THE KEY and alluded to it earlier because if government is not ready for this technology to be used when it becomes available then there will a swath of differing technologies being used on the road by early adopters regardless of the law.
A framework must be developed at a federal level to standardise the technology and laws across the nation.

Lawyers. See above. If the law is not clear and unambiguous then it becomes open to legal interpretation and lawyers will have a field day. This in itself could become a huge impediment to a faster uptake and development of the technology. If car makers are reluctant don't have the protection of law and they find themselves being sued if their product is misused then they could be very reluctant to bring their products to market without a lot of locks and controls.

Software testing and control. I addressed this in my previous blog regarding hacking and it is a serious concern.

Improved sensors. The current technology appears to have reached the point of being practical and usable but needs to be extended for distance and weather conditions as they can become inaccurate and obscured by rain, snow and fog.  
 However this shortcoming can be overcome to a large extent by detailed GPS based mapping so that if the sensor lose real time visibility it could slow the vehicle and continue to run by utilising GPS coordinates and comparison to the stored visual data. This detailed mapping is being done now by the likes of Google.

Vehicle to vehicle and infrastructure communication. To allow a vehicle to know what is happening and to share information about road conditions, weather, roadworks or collisions and natural disasters it is essential to have real time communications between vehicles and infrastructure. This allows the cars to tell each other not only about conditions but also about what actions they intend to take in a given situation, ie emergency braking or avoidance. This needs to be standardised and implemented, but again, government has to have a major role in ensuring this works for interoperability between manufacturers.

A case could be made to fit ALL existing vehicles with this so autonomous cars know what manual vehicles are doing as well as alerting their drivers to dangerous conditions that other vehicles have reported. China is currently rolling out this before the autonomous cars hit the road in a proactive move, but again there could be other considerations to that in a totalitarian state.

Strangely enough, each time I look at the progress of automated vehicles I keep coming back to the same point.

We can't leave the development of such an important and social changing technology to the manufacturers alone.

Governments need to be involved NOW before its too late, at least in a standards framework design by manufactures and other stakeholders to come up with something everyone can live with (literally).

Autonomous vehicles in the new world - Who is driving? seems to be an appropriate sentiment.

Part 9 - Highway to hell or paradise by the dashboard light?

Autonomous vehicles, they are coming, no question, no way to get the genie back in the bottle, autonomous vehicles will be a part of everyday life. Without doubt they will have a huge impact on life, transport, and anything else you can think of but there is also a very great potential for a very dark side.

Anyone who has had an even passing use of a computer at home or at work knows about the minefield individuals and companies need to go through every single day to keep them secure, virus free and usable. 

When a hacker gets into your computer it can have a devastating effect on your life ranging from stealing your bank accounts right up to pretty much stealing your entire life with identity theft.

Now consider that every function of an autonomous car is controlled by one or more computers. It's not such a big step to imagine what would happen if a freeway full of autonomous cars travelling at high speed suddenly got hacked? Identity theft would look pretty good in comparison to the widespread death and destruction that could be the result of a state sponsored attack on the road system of another country.

I don't for a minute doubt that this will be addressed before a widespread release of this technology happens, but as history shows us, we don't know what we don't know.

For example if you own a Jeep Cherokee there is a chance that someone could have taken control of your car and crashed it for you. 

Wired magazine has very good story here about two researchers who did just that. Charlie Miller and Chris Valasek have found (and exploited) a weakness in Chryslers' systems that allowed remote access once they have the IP address of a computer installed in these cars called Uconnect.

Once they had that they could download code that effectively gave them control of everything electronic within the vehicle..... including the steering and brakes.

To be fair this never made it into the wild as Chrysler issued a fix extremely quickly.

But as I said, we don't know what we don't know so the possibility of someone with evil intent finding a flaw like this and exploiting it for nefarious purposes is a very real possibility.  

To me, this is another reason that governments need to be proactive in this technology, not by controlling it but setting the laws and standards of testing to ensure that issues such as this can be mitigated.

Not that industry would deliberately falsify or compromise safety standards or emission controls would they? OK Volkswagen springs to mind with diesels, oh, and with cars stopping for no reason, and yea, Toyota cars failing to stop and... you get the idea. Admittedly the Volkswagen and Toyota problems with stopping and failing to stop respectively appear more to do with software testing than malevolence but just goes to prove the point that we don't know what we don't know. The manufacturers thought that the software that controlled these functions worked just fine but they didn't know that certain circumstances could create problems.

I don't claim to know the inner workings of car design but it needs oversight in the form of regulation to ensure that testing is rigorous and effective and as the car assumes more and more of the driving functionality it becomes more and more critical that the software is up to the job.

Another software control that needs to be built in and I know that some cars already have this, is multiple computers that cross check each other, so if there is a failure, the cars knows it's ill and can safely move off the road while refusing to go back into autonomous mode. 

I'm sorry Dave, I'm afraid I can't do that!

All of the onus isn't on the car manufacturers either. As this technology becomes more pervasive the road system will need to become smart as well. No longer will it be enough to have speed limit signs and road work signs. Some cars, such as Mercedes can read these but it would be far better to make the road system more intelligent and proactive. That way the car can know road conditions and delays before it gets there and if there is a more efficient route, change its course. Road works with temporary lane markings and speed limits can be transmitted electronically (and enforced).

The software that controls that will need to be vigorously tested and secured as well, as there are public safety implications to be protected.

I never thought that I would be an advocate for government to be involved in private industry decisions and I still don't but I do believe there needs to be a government framework around all of this to ensure that public safety comes before the corporate need. 

Corporations are there to make profits for their shareholders and sometimes the corporation comes before the welfare of the people. 

For any large organisation, be it government or corporate, there has to be a system of checks and balances for the greater good.  

I keep saying, and will keep saying that government needs to be involve now, before the car makers set the standards and build a power base that can't be reversed.

The time to do this is now, before it becomes mainstream, by then it may be too late and any changes or legislation once there is a large user base will only serve to delay the implementation of this technology, not assist in its uptake.

Now for something completely different.

New technology of the week, only from the Dutch.