Automotive Technology Blog

Automotive Technology

The Automotive Technology Blog is the place for conversation and discussion about electrical/electronic components, materials, design & assembly, and powertrain systems. Here, you'll find everything from application ideas, to news and industry trends, to hot topics and cutting edge innovations.

Previous in Blog: NYC Orders Solar Powered EV Charging Stations  
Close
Close
Close
13 comments

Algorithms Under Control: The Efforts to Help Self-driving Cars Remain Ethical via the Trolley Problem

Posted February 28, 2018 1:37 PM by IronWoman

I think by now we’ve all heard the hype surrounding automated vehicles. Sure, they seem like everything people have been hoping for since the airing of The Jetsons, but ask yourself one question: are they ethical? While it isn’t necessarily the first thought that comes to mind when imagining self-driving cars, a handful of professionals have paid it some heed. Olivia Goldhill explains further in her article “Philosophers are Building Ethical Algorithms to Help Control Self-driving Cars.”

Goldhill’s piece, published by Quartz Media LLC, delves into the potential downfalls of pre-programmed automobiles, or what philosophers are now calling a “myriad of ethical quandaries on wheels.” Their problem is with the Enactment of Moral Conundrum aka the Trolley Problem. To explain further, the Trolley Problem is a thought experiment in ethics: there is a runaway trolley racing down the tracks, headed straight for five people tied up on said tracks and unable to move. If you pull a lever, the trolley will switch to a different set of tracks with one person tied up on the side track. You can either do nothing, killing the five people, or pull the lever, diverting the trolley onto the side track where it will kill one person.

So which is the most ethical choice? Nicholas Evans, philosophy professor at UMass Lowell, headed a study alongside two other philosophers and an engineer, where they built algorithms for various ethical theories – a language that can be read by computers. Their hope is that cars may have the ability to be programmed, when the time comes, to make specific moral split-second decisions. According to Goldhill, “[t]heir work, supported by a $556,000 grant from the National Science Foundation, will allow them to create various Trolley Problem scenarios, and show how an autonomous car would respond according to the ethical theory it follows.”

Overall, the object of Evans and his team’s study is to garner results that influence others – like consumers or manufacturers – to make an informed decision. As Evans said, “…we have to have a discussion as a society about not just how much risk we’re willing to take but who we’re willing to expose to risk.” While the group hopes to eventually collaborate with autonomous car companies on this issue, other topics on the table include how to build traffic infrastructure to accommodate the advancements of personal and commercial vehicles.

An important question that arose during Evans and his cohorts’ study was: how could a set of values used to program self-driving cars be hacked (i.e. through physical changes like installing weaponry, or making the vehicle volatile in its decisions)? In addition, say manufacturers are able to pre-program autonomous vehicles specific to the needs and preferences of its driver. How will differently programmed vehicles react with one another on the road? In a separate yet related study, psychologists have begun research on the topic of autonomous cars and continued problems, conducting polls to uncover which moral solution of the vehicle’s choosing is the majority favorite among consumers.

At the end of the day, Nicholas Evans understands that his research only scratches the surface, not addressing the broader ethical issue surrounding autonomous cars. However, Cal Poly San Luis Obispo philosophy professor Patrick Lin has delved into it on his own, testing ethics encompassing advertising, liability, social issues and privacy when it comes to autonomous vehicles without incorporating the Trolley Problem. Another big question surrounding this change in travel has to do with survival rates: what hypothetical negative results will transpire from the positive results of people surviving car crashes? Will we suffer from rapid overpopulation, survivor’s guilt, etc.? The world may never truly know until we reach that point.

After reviewing the conducted research thus far, it is safe to say that autonomous cars will create massive unforeseen effects. I suppose that, for now, we’ll have to keep our eyes and ears open for studies to come – hopefully being mindful of advances towards autonomous car safety. Do you have any predictions when it comes to driverless technology in our future? Are you optimistic about this technological development or do you view it as a misstep?

Sources: Quartz Media, LLC, XKCD, Watt Knowledge, Monday Note

Reply

Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Guru

Join Date: Apr 2010
Location: About 4000 miles from the center of the earth (+/-100 mi)
Posts: 9146
Good Answers: 1038
#1

Re: Algorithms Under Control

02/28/2018 3:37 PM

I could be wrong, but I'm thinking that the first time an autonomous car runs over a child chasing a ball into the street from behind a parked car, there will be a big public outcry. People need someone to be responsible. I suspect the lawyers will be circling around looking for someone to sue.

Personally, I don't like the idea of a machine deciding how to kill the fewest people, but on the other hand, I feel the same way about some of the people on the road, yapping on the cellphones and not paying attention to their driving.

Reply
Guru
Engineering Fields - Control Engineering - Time to take control United States - Member - New Member Engineering Fields - Systems Engineering - New Member Engineering Fields - Mechanical Engineering - New Member

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
Good Answers: 87
#2

Algorithms Under Control: The Efforts to Help Self-driving Cars Remain Ethical v

03/01/2018 6:48 AM

It seems to me that the Trolley problem as stated here is a no brainer. How could allowing 5 people to die be more ethical than the allowing 1?

If one argues that 'allowing' is different than 'taking action'. My position on that is that by doing nothing, that is itself an ethical choice.

I think a Trolley problem that better challenges ethics would be where that 'one' person is a family member or someone highly important in society. Then the problem is do nothing causing 5 deaths and saving someone important to you or take action to save 4 lives (saving 5 from death and causing 1 death).

__________________
J B
Reply
Power-User

Join Date: Nov 2009
Posts: 413
Good Answers: 23
#3

Re: Algorithms Under Control

03/01/2018 7:46 AM

The simplest and cheapest way to achieve control with safety in the long run is going to be installing control systems embedded in the roads that control the spacing and the speed of all vehicles, and every vehicle that wants to travel those roads will need to be forced to carry the universal, standard issue limiters that connect those vehicles to the inlaid control system, or they simply won't be allowed entry on to that road.

Just like railroads that operate with perfect computer control.

That way the control is taken out of the hands of the bird brains in Con valley, like Tesla, that in their retarded logic put holding down costs before their customers' safety which consequently led them to believe they were entitled to put an unguided projectile on the public roads that didn't even have the ability to see a solid blockage directly in front of it because the projectile was in the words of the self appointed "Masters Of The Universe", in its "beta" phase.

"Beta" testing is how the bird brains in Con valley get the dim witted soporific public to test their death traps for them without any liability falling back on the bird brains.

Read the terms of use on any Con valley product and you will see that you have no comeback whatsoever, whatever they want to impose on you.

Inlaid controls in the road will defeat the corner cutting efforts of the bird brains who will be neutered in the name of a compulsory adherence to a universal safety system.

More than 20 years after the widespread adoption of desktop computers the Con valley bird brains cannot even get their latest desktop computers to operate 100% perfectly and assigning your safety to them by giving them a free pass to put their hacks on moving lumps of metal is to vacate your logic and reason so completely that you may as well drink a gallon of hard liquor before getting in one of their projectiles because that will at least keep you blotted out from reality while the bird brains remove your head from your shoulders in a most unsurgical way.

Reply
Guru

Join Date: Oct 2009
Posts: 1449
Good Answers: 30
#7
In reply to #3

Re: Algorithms Under Control

03/01/2018 5:28 PM

I wonder what you are going to say when you discover the control systems in the road have been embedded as part of their beta-testing

Reply
Guru

Join Date: Jul 2012
Location: Madison, WI.
Posts: 2093
Good Answers: 80
#4

Algorithms Under Control: The Efforts to Help Self-driving Cars Remain Ethical v

03/01/2018 11:57 AM

This is not as big a problem as the nervous Nellys would like you to believe.

Humans are made to face that test daily and more often than not fail to make the ethical choice because of that "third force" of the universe, Enlightened Self-interest.

Meanwhile,

The "researchers" are all lining up for government grants to "study" a non existent problem.

The other camp are the lawyers who are salivating uncontrollably at the possibilities for multi million dollar pay days.

In the middle are people like me who drive a fair amount and can tell you without reservation that by taking the wheel out of peoples hands the roads will be safer, but only when all cars on that road are fully auto and connected.

Inevitably, in 50 years or so, private vehicle ownership will be the province of the rich only. The same rich who are renting you rides in their fleet of autonomous vehicles.

__________________
Knowing is the end result of learning, not believing.
Reply
Guru
Hobbies - CNC - New Member Popular Science - Biology - New Member Hobbies - Musician - New Member

Join Date: Dec 2008
Location: Canada
Posts: 3499
Good Answers: 145
#5

self driving cars

03/01/2018 1:28 PM

If the speed limits on highways hadn't been bumped up to simply unsafe speeds we wouldn't need autonomous (or inter-connected/controlled by one means or another) vehicles.

If as much effort was put into high quality public transportation, once again, absolutely no need for autonomous private vehicles (except for a few rich lushes who are too cheap to pay a chauffeur).

In some parts of the world, even though we have stupidly high speed limits and inadequate or non existent public transport, I can't see autonomous vehicles being billed as safe, where the real safety issues including moose, ice, crazy windstorms, snow, etc are complex and randomly inflicted upon the travelling vehicle no matter who or what is driving. "Other drivers" sure they add a few risks but the overwhelming majority of risks are not driver-related. Autonomous cars would have nearly as many accidents IMO, only drop a few due to phone calls and text messages, old driver heart attacks etc.

Also, for those of us who prefer to be the driver and rarely enjoy being the passenger, what kind of "back seat driver" interface is the company going to provide for us in our "driverless" car? Can you tell me we'll be able to enjoy the ride never wanting to yell at or at least pointedly criticize the performance of the thing? I don't believe it. Can you tell me this automaton will not be utterly narcissistic about its ability to drive? In short, the most annoying kind one can ever be a passenger to? If they can come up with an AI that satisfies my need for a proper response to 'back seat driving' when called for, then I might consider a ride in a driverless car.

__________________
incus opella
Reply
Guru

Join Date: Apr 2010
Location: About 4000 miles from the center of the earth (+/-100 mi)
Posts: 9146
Good Answers: 1038
#6

Re: Algorithms

03/01/2018 5:18 PM

It doesn't matter to me, I'm not buying one. I can drive myself...

Reply
Guru
Engineering Fields - Optical Engineering - Member Engineering Fields - Engineering Physics - Member Engineering Fields - Systems Engineering - Member

Join Date: Apr 2010
Location: Trantor
Posts: 5363
Good Answers: 646
#8

Trolley Problem

03/02/2018 2:58 PM

When a problem is stated that has an inadequate premise, it's no surprise that the answer is unsatisfactory.

The premise is that autonomous vehicles (AVs) will have the same braking systems that ordinary cars do (or, trolleys). And yes, they do at the present time, but there are a number of other braking mechanisms that can be deployed to stop a car in a much shorter distance if modifications were made to the undercarriage. Right now, cars use the friction created by small brake pads and the contact area of the tires to stop a car.

However, if a large friction plate (or, drag plate) was fitted under the car that occupied the entire area between the 4 wheels, and this friction plate was dropped abruptly in an emergency stop situation to make contact with the road, the force of friction would be greatly magnified and the car could almost literally 'stop on a dime'. Airbags within the passenger compartment would likewise be instantaneously deployed to protect the passengers.

In fact, the system could be refined to have left and right frictions plate under computer-navigation control, so that by carefully applying pressure to the friction plates, the car could be 'steered' to avoid inadvertent drifting.

__________________
Whiskey, women -- and astrophysics. Because sometimes a problem can't be solved with just whiskey and women.
Reply
Guru
Engineering Fields - Control Engineering - Time to take control United States - Member - New Member Engineering Fields - Systems Engineering - New Member Engineering Fields - Mechanical Engineering - New Member

Join Date: Apr 2009
Location: Tampa, Florida, USA
Posts: 2105
Good Answers: 87
#9
In reply to #8

Re: Trolley Problem

03/02/2018 3:30 PM

I'm curious, if it's possible to have such a breaking mechanism now and such a mechanism would make the highways safer, then why are they being used now?

What difference does it make if the car is autonomous or not?

__________________
J B
Reply
Guru

Join Date: Apr 2010
Location: About 4000 miles from the center of the earth (+/-100 mi)
Posts: 9146
Good Answers: 1038
#10
In reply to #8

Re: Trolley Problem

03/02/2018 10:13 PM

I'm thinking that the braking mechanism you describe might cause more problems than it would solve. Stopping very quickly would either cause cars behind you to collide with your vehicle or to also brake very quickly and the likely result would be a massive pile up.

The present system allows for controlled braking to avoid an obstacle in front while not becoming an obstacle for the following driver.

Reply
Guru

Join Date: Oct 2009
Posts: 1449
Good Answers: 30
#11
In reply to #10

Re: Trolley Problem

03/03/2018 7:33 AM

Not to mention the damage caused to the road surface.

Reply
Guru

Join Date: Apr 2010
Location: About 4000 miles from the center of the earth (+/-100 mi)
Posts: 9146
Good Answers: 1038
#12

Unethical autonomous cars

03/03/2018 3:21 PM

Unethical autonomous vehicles should be sentenced to serve time in the impound lot...

Reply
Power-User

Join Date: Nov 2009
Posts: 413
Good Answers: 23
#13

Re: Algorithms Under Control:

03/06/2018 9:52 AM

Google executive Eric Schmidt on the subject:

“Let me be clear: Humans will remain in charge of [AI] for the rest of time. The other point that I want to remind everyone, these technologies have serious errors in them, and they should not be used with life-critical decisions. So I would not want to be in an airplane where the computer was making all the general intelligence decisions about flying it. The technology is just not reliable enough. There too many errors in its use. It is advisory, it makes you smarter and so forth, but I wouldn’t put it in charge of command and control.”

Reply
Reply to Blog Entry 13 comments
Interested in this topic? By joining CR4 you can "subscribe" to
this discussion and receive notification when new comments are added.
Copy to Clipboard

Users who posted comments:

artsmith (1); JBTardis (2); phph001 (2); rashavarek (1); Rixter (4); Usbport (1); wrenchtwirler (2)

Previous in Blog: NYC Orders Solar Powered EV Charging Stations  

Advertisement