Autonomous Cars: Who Gets Protected?

[Elias]

Hi All,

I thought the below article would be a good conversation starter! A few thoughts spring to mind including, can it be designed better to prevent the situation entirely?

Driverless Cars: Who Gets Protected?

Regards,

Elias

[Geoff Sizer]

I am personally sceptical about the prospects of driverless cars gaining acceptance for widespread use, at least in the short to medium term.  I would certainly be uncomfortable about committing my life to one as a passenger, and pale at the thought of having driverless cars around me on the road.

Apart from decision-making when faced with irreconcilable conflicts, what about the competence and ethics of the programmers responsible for implementing such choices?  Recent experiences with the VW emissions scandal and uncontrolled acceleration in Toyotas gives one little cause for comfort.

Perhaps we should step back and follow the path of driver augmentation systems which are emerging, and see how they pan out before taking humans out of the decision-making loop.

[Geoff Sizer]

This link touches on the insurance implications of driverless vehicles:  http://prwire.com.au/pr/58154/internet-of-things-and-increasing-digital-demand-signals-major-insurance-industry-disruption

[Justin Blows]

I often wonder why aircraft are so safe.  Is it because of the safety protocols that are strictly followed by pilots?

There is a potential to enforce safety protocols in driverless cars that the human population will not follow, which may make transportation much safer.

Something to aim for, at least.

[Justin Blows]

I just spotted this article about a raspberry PI AI beating a fighter pilot - every time.  If a PI can do this for dog fights, image what is possible for cars.

[Ankur Barua]

Congratulations to the group on a great conversation starter.

If we look at the evolution of of the automated vehicle (say using US based NHTSA system of 1 to 4) and their application in different environments it is possible that the choices do not need to be mutually exclusive. For example a level 1 vehicle with Auto Emergency Braking, Adaptive Cruise Control and Electronic Stability Control would appear to be great fit for our roads and driving conditions for now. Similarly there are certain cities elsewhere in the world where traffic congestion and driving conditions are such that a level 4 vehicle (fully automated) may actually improve quality of life for numerous people. On the subject of quality of life what does it do for people with limited mobility?

The point of the above is steady progress is possible without the moral quandary if the context is suitable and the assessed benefits outweigh the risks. Since we are about IoT it will be great to get some views on Connected Vehicles as well which complement Automated Vehicles and will be a determining factor in applications such as public transportation.    

[Justin Blows]

That's a great way to look at the problems, Ankur.  I remember being driven in Vietnam.  They could really do with some sort of machine organisation and control of vehicles on the road.  It's chaos their.  I suspect that even with less than perfect systems safety would be dramatically improved.

[Justin Blows]

Very sad - The Guardian is reporting the first recorded death because of an autonomous car failure.

 

Edited June 30, 2016 by Justin Blows

[Raj Purushothaman]

Is lot scope for IoT in the Rail Industry? There are existing sensors are wired to panels ( Microlok etc) for status of trains; is this status could be changed to wireless for railway IoT?

There are semi- autonomous trains exist in Europe with attendant / driver in control in case of emergencies. Scope for fully autonomous train in the future through IoT applications??

 

Regards

Raj Purushothaman

[Ben Macey]

I think hard wired control systems will remain in use in the rail industry. Safety control systems typically require fast signals which will not be compromised by wireless communications. Typically they are also devices which require power for the inputs and power for the outputs so hard wiring is the best option.

I can imagine automation increasing in the rail industry. I cannot imagine driverless trains in the short term.

[Chi Binh Le]

On 05/07/2016 at 0:44 PM, Ben Macey said:

I can imagine automation increasing in the rail industry. I cannot imagine driverless trains in the short term.

 

I would have thought that driverless trains would be an ideal starting point for driverless vehicles in general. There are several advantages to trains:

1. They run on well defined tracks

2. There are fewer trains in a rail network than cars in the road network which simplifies the infrastructure/vehicle and inter-vehicle communications

3. There is already a standard for rail network communications (GSM-R) which could be extended to handle more situations

4. They don't have to contend with pedestrians

[Ben Macey]

Chi Binh Le,

I agree, I view the challenges in the acceptance by unions and safety concerns form users and pedestrians at crossings.