Future of transport

[TechMonkey]

Or the human pilot who deliberately kills everyone on the plane. I'd rather the computer flew the plane than the sleep deprived depressed alcoholics we seem to get too often

Only time the DLR crashed is when humans were driving

 

Part of the issue is hyperbole like this though. Flying is still counted as the safest form of transport, so saying all the pilots are sleep deprived, depressed, alcoholics is just provably incorrect and makes any argument it is attached to look silly.

Your DLR example is also silly as there have been two crashes. One was a software issue they were testing, so actually goes against your point; the other was between 2 trains, one automated, one manual. I'd count the latter as a no score draw. Using the DLR as an example would bring up an incident where someone fell in front of one of the trams and due to no driver didn't stop, though there would have been time and space to do so.

 

Using facts and figures are a much better way than silly snarky comments as all they do is distract from the matter at hand.

[mavhc]

Who said all pilots, I said too often pilots. But it's true that plane crashes get attention because they're rare, vs car crashes that don't get attention because they're common.

 

On 10 March 1987, before the system opened, a test train crashed through buffer stops at the original high-level Island Gardens terminus and was left hanging from the end of the elevated track. The accident has largely been attributed to unauthorised tests being run before the correct installation of the wayside safety system had been verified; an omission in the wayside system allowed the train to travel too fast on the approach to the terminus. The train was being driven manually at the time.

 

On 22 April 1991, two trains collided at a junction on the West India Quay bridge during morning rush hour, requiring a shutdown of the system and evacuation of passengers by ladder.[182][183] One train was travelling automatically; the other was under manual control

 

On 9 February 1996, the Provisional IRA blew up a lorry under a bridge near South Quay,[185] killing two people and injuring many others.[186] This caused £85 million of damage and marked an end to an IRA ceasefire in force at the time. Significant disruption was caused and a train was stranded at Island Gardens, unable to move until the track was rebuilt.

[mavhc]

https://waymo.com/research/do-autonomous-vehicles-outperform-latest-generation-human-driven-vehicles-25-million-miles/

[6Foot2]

https://waymo.com/research/do-autonomous-vehicles-outperform-latest-generation-human-driven-vehicles-25-million-miles/

 

To support your position on autonomous/driverless cars, you put forward publicity pushed by Waymo - an autonomous driving technology company.

 

Any information coming from them is unlikley to show much about any potential negative implications of driverless cars.

[TechMonkey]

Who said all pilots, I said too often pilots. But it's true that plane crashes get attention because they're rare, vs car crashes that don't get attention because they're common.

 

On 10 March 1987, before the system opened, a test train crashed through buffer stops at the original high-level Island Gardens terminus and was left hanging from the end of the elevated track. The accident has largely been attributed to unauthorised tests being run before the correct installation of the wayside safety system had been verified; an omission in the wayside system allowed the train to travel too fast on the approach to the terminus. The train was being driven manually at the time.

 

On 22 April 1991, two trains collided at a junction on the West India Quay bridge during morning rush hour, requiring a shutdown of the system and evacuation of passengers by ladder.[182][183] One train was travelling automatically; the other was under manual control

 

On 9 February 1996, the Provisional IRA blew up a lorry under a bridge near South Quay,[185] killing two people and injuring many others.[186] This caused £85 million of damage and marked an end to an IRA ceasefire in force at the time. Significant disruption was caused and a train was stranded at Island Gardens, unable to move until the track was rebuilt.

 

You said too often, makes it sound endemic. In 2022 there were 7 reported alcohol violations in the US, for around 16.4 million departures.

 

If you didn't selectively copy and paste it also helps. From the same article you copied from regarding the 10th March 1987:

 

However, inside sources have stated these tests were being done to test the ATP if the train was entering the station too quickly, and after six successful tests, a software issue involving asynchronous computer systems caused the train to not receive the instructions to activate the ATP, cut off power from the motors, and apply the brakes soon enough to prevent the train from slowing down, causing the train to go through the buffers. Following this, the software was reworked and the braking distance was changed to ensure such incidents wouldn't happen in practice, and as a result may have possibly prevented a far more tragic incident from occurring.

DLR Wikipedia

 

So testing for a software fault that would have caused a worse incident happening.

 

Beyond that I'm not sure the point of your post, you are just repeating what I said. There has been more than one crash on the DLR and both have involved automated systems.

[mavhc]

I mean, that's why you test things before they go into production. Can't complain they found a bug before letting passengers on.

 

On planes: Royal Air Maroc Flight 630, SilkAir Flight 185, EgyptAir Flight 990, LAM Mozambique Airlines Flight 470, Malaysia Airlines Flight 370, Germanwings Flight 9525, China Eastern Airlines Flight 5735 all crashed on purpose by the pilot.

 

Waymo has driven 25 million miles and killed no one

[LeMarchand]

 

Just sayin'

[mavhc]

the dangers of cgi cars is not to be overstated

[NegativeKillDeath]

There are only 1.27 deaths per 100 millions miles driven in the US in total so the Waymo statistic isn't a good indicator. Probably why their reports focuses on insurance claims for property and bodily injury claims.

[mavhc]

To support your position on autonomous/driverless cars, you put forward publicity pushed by Waymo - an autonomous driving technology company.

 

Any information coming from them is unlikley to show much about any potential negative implications of driverless cars.

 

Or I linked to a peer reviewed scientific paper lead by a 161 year old insurance company with $180 billion in assets, who can tell

[TechMonkey]

I mean, that's why you test things before they go into production. Can't complain they found a bug before letting passengers on.

 

On planes: Royal Air Maroc Flight 630, SilkAir Flight 185, EgyptAir Flight 990, LAM Mozambique Airlines Flight 470, Malaysia Airlines Flight 370, Germanwings Flight 9525, China Eastern Airlines Flight 5735 all crashed on purpose by the pilot.

 

Waymo has driven 25 million miles and killed no one

 

Which was what they were doing. So, I am confused at your point. You've gone from the only crash on the DLR was human error to yes the crash was valid as they were testing?

 

What time frame were those crashes over?

 

& yet we have an Uber death & forty-eight confirmed Tesla fatalities.

[mavhc]

Auto pilot isn't self driving though, it's ADAS. No one said you should let it drive itself. In fact no one said you should let FSD drive itself, and 2 people have died while it's been in use, which is pretty good for 2 billion miles

 

There's 3 crashes on the DLR, one from testing, one from a human driving, and one from a bomb. None from the computer controlled system in production.

 

Of course it does keep running into people because it has no cameras, they should fix that

[6Foot2]

Link: Iowa City Made Its Buses Free. Traffic Cleared, and So Did the Air. - The New York Times

 

Quote

There was a psychiatrist, a librarian, a substitute teacher and a graduate student in biomedical engineering. There was an Amazon warehouse worker who’d just finished his night shift, and a man who’d lost his driver’s license because of an incident in Florida that he didn’t want to talk about.

 

They were all riding Iowa City’s buses one sunny November morning, and they were all amped about the same thing: That everyone got to ride for free.

 

Iowa City eliminated bus fares in August 2023 with a goal of lowering emissions from cars and encouraging people to take public transit. The two-year pilot program proved so popular that the City Council voted this summer to extend it another year, paying for it with a 1 percent increase in utility taxes and by doubling most public parking rates to $2 from $1.

 

Ridership has surpassed prepandemic levels by 18 percent. Bus drivers say they’re navigating less congested streets. People drove 1.8 million fewer miles on city streets, according to government calculations, and emissions dropped by 24,000 metric tons of carbon dioxide a year. That’s the equivalent of taking 5,200 vehicles off the roads...

 

[mavhc]

Updated FSD collision stats https://www.tesla.com/en_gb/fsd/safety

 

Also new internal messages indicate FSD coming to some parts of Europe soon

 

FSD in UK 

 

[6Foot2]

Link: EU waters down plans to end new petrol and diesel car sales by 2035 - BBC News

 

Quote

The European Union (EU) has watered down its plans to ban the sale of new petrol and diesel vehicles by 2035.

 

Current rules state that new vehicles sold from that date should be "zero emission", but carmakers, particularly in Germany, have lobbied heavily for concessions.

 

Under the European Commission's new plan, 90% of new cars sold from 2035 would have to be zero-emission, rather than 100%.

 

According to the European [carmakers'] association, ACEA, market demand for electric cars is currently too low, and without a change to the rules, manufacturers would risk "multi-billion euro" penalties.

 

The remaining 10% could be made up of conventional petrol or diesel cars, along with hybrids.

 

Carmakers will be expected to use low-carbon steel made in the EU in the vehicles they produce.

 

The Commission also expects an increase in the use of biofuels and so-called e-fuels, which are synthesised from captured carbon dioxide, to compensate for the extra emissions created by petrol and diesel vehicles...

 

[mavhc]

Water's up then.

 

In good news cheap self driving cars are here, so there will be little reason to buy a car at all soon

[Ditto]

When is soon?

[sigma]

“How soon is now?”

[mavhc]

I'll bet on 5 years, specifically by 1st Jan 2031 self driving taxis will be everywhere, and cheaper to use than owning a car, if you travel less than 20000 miles per year, not including flights over 500 miles

[Ditto]

I suspect you underestimate the resistance to change from the general public. Taxi's could provide the service you describe and would be more cost effect than the 20000 miles travel and car ownership. Actually, given taxi prices around here, maybe much less! The vigour with which large numbers of car owners value the ability to get up and go at any point in time, regardless of cost, is something not to be underestimated. Perhaps you could diary a revist in 5 years time, and see which of have the stronger Nostradamus attributes!

[dcwhitworth]

2 minutes ago, Ditto said:

I suspect you underestimate the resistance to change from the general public. Taxi's could provide the service you describe and would be more cost effect than the 20000 miles travel and car ownership. Actually, given taxi prices around here, maybe much less! The vigour with which large numbers of car owners value the ability to get up and go at any point in time, regardless of cost, is something not to be underestimated. Perhaps you could diary a revist in 5 years time, and see which of have the stronger Nostradamus attributes!

 

Yes the illogicality of the financial side of car ownership is going to be hard to overcome. And if there's lots of money going somewhere you can be sure there will be powerful lobbies trying to keep it that way, coupled with the inertia/short termism/cowardice of politicians and the general public distrust of technology, I fear five years would be highly optimistic.

[mikes]

I highly doubt workable self driving cars/taxis will be in the UK in 2030,  try 2040.  They have fallen foul to the "90% of the task is very easy, but the last 10% is very hard" problem - in the early days they were making a lot of progress and were very optimistic (Elon saying in 2015 that by 2020 you'd be able to buy a Tesla to make money for you loaning it out as a taxi).   

 

What progress have they made in the past 5 years ?  Not much and in fact a lot of these self driving car firms now either have very restricted areas they must manually keep upto date - doesn't scale at all nationwide - or they have a team of humans who have to manually take control of cars the 5% of the time they get stuck - again not scalable at all.

 

Now maybe if a cheap LIDAR system becomes available and the car makers can implement that well along with the existing video only system Tesla uses...  but I would not feel comfortable at all falling asleep in a video camera only self driving system, hell no, not in 5 years time either

[LeMarchand]

On 05/01/2026 at 10:15, mavhc said:

I'll bet on 5 years, specifically by 1st Jan 2031 self driving taxis will be everywhere, and cheaper to use than owning a car, if you travel less than 20000 miles per year, not including flights over 500 miles

 

Apart from the tech problems, I can't see the car manufacturers liking that idea. Also the subsidiary businesses: parts, repairs, accessories etc. What about families who need child seats etc? Interior cleanliness?

[mavhc]

They're already here and people love them

 

FSD 7 billion miles https://www.tesla.com/en_gb/fsd/safety

FSD Robotaxi 250,000 miles

Waymo 127 million miles https://waymo.com/safety/impact

 

Cheaper does not mean everyone will switch, people will pay incredible amounts to not have to put a child seat in a car every time (or just request one with a child seat)

 

But the main reason to do it is humans kill 3000 people a day while driving

 

https://injuryfacts.nsc.org/all-injuries/costs/guide-to-calculating-costs/data-details/ $2 million per death, so 2 trillion per year

[dcwhitworth]

38 minutes ago, mavhc said:

They're already here and people love them

 

FSD 7 billion miles https://www.tesla.com/en_gb/fsd/safety

FSD Robotaxi 250,000 miles

Waymo 127 million miles https://waymo.com/safety/impact

 

Cheaper does not mean everyone will switch, people will pay incredible amounts to not have to put a child seat in a car every time (or just request one with a child seat)

 

But the main reason to do it is humans kill 3000 people a day while driving

 

https://injuryfacts.nsc.org/all-injuries/costs/guide-to-calculating-costs/data-details/ $2 million per death, so 2 trillion per year

Yes, they don’t need to be perfect, just significantly safer than humans and that’s a pretty low bar. A major legislative hurdle is liability, if a self driving car causes an accident who is liable ? The owner or the manufacturer ?

 

Another likely stumbling block is the (to my mind) weird and entitled attitude of many car drivers in this country. The outrage that abounds when someone tries to hold them to account for speeding or illegal parking, or otherwise dares to try and restrict them, is head scratching to me. I find it similar to the American obsession with guns.

Edited January 9 by dcwhitworth