GM's Super Cruise: Driving a 2018 Cadillac CT6 with a Full-On Auto Pilot | Hacker News Reader

GM's Super Cruise: Driving a 2018 Cadillac CT6 with a Full-On Auto Pilot

motortrend.com

33 pointsedward58 comments

Given all the restrictions it sounds like the same old adaptive cruise control and lane assist that has been around for years. Actually less appealing, with all the face monitoring and eye tracking, not to mention that it only works on 'vetted' stretches of road.

Also, I wouldn't be caught dead in a car that looked like that, but that's another matter.

> Given all the restrictions it sounds like the same old adaptive cruise control and lane assist that has been around for years. Actually less appealing, with all the face monitoring and eye tracking, not to mention that it only works on 'vetted' stretches of road.

Yes that sounds like quite the restrictions. The vetted road one is interesting as in theory they could expand it with post purchase updates (or even over the air). But given that these are the same companies that would charge $250+ for a DVD of new maps for a built in GPS nagivation system ... well you get the idea.

> Also, I wouldn't be caught dead in a car that looked like that, but that's another matter.

It's a subjective topic but I think it looks quite nice. Seems halfway between the previous Caddy body style and a more German styling.

The road restriction is to all limited access highways in the US and Canada. It's less of a vetting and more related to the type of driving they are willing to turn the system on for.

I think it's more limited than that as it refers to "meticulously lidar-mapped." But no idea how limited or not that restriction is.

They announced it at the time, they LIDAR mapped all of the limited access roads in the US and Canada.

http://media.gm.com/media/us/en/cadillac/news.detail.html/co...

I'd be super curious to know why exactly American cars look so unappealing to people on the other side of the pond. I find them bloated, full of plastic, tacky, tastless, the lights on front are always this elongated very unattractive shape - and yet, it must be what Americans like, since these are cars designed by them. I wouldn't be surprised if they found European cars ugly in return - Is there any study done on this? An objective look into why Americans/Europeans like what they do in cars?

I'm curious about this as well. I know it's a bit taboo to say here but I don't think the Tesla S looks very attractive either. Something about the shape and proportions that makes it look bulbous and just a bit amateurish. I'm also European and used to European design language.

I think my first reaction of seeing a Tesla Model S was "that's a nice looking Mondeo". I actually think they are quite handsome in an understated kind of way.

However, my first reaction when seeing a (white) Model X was "that's incredibly ugly".

NB I am in the UK

Does anyone find the X pretty? It's a godawful abomination of a vehicle.

The thing that gets me about the X is this. While I know cross-overs are mostly a design statement rather than being about actually living the wild sporting life, the fact that you have something that's nominally an SUV that can't have a roof rack because of the gull wing doors strikes me as pretty silly. (Of course, EVs aren't great for lugging canoes to remote locations anyway.)

Coming from the UK, I've noticed this too - especially with interiors. For example, I don't think the exterior of the GM Bolt looks too bad, but the interior just looks so cheap [1] and that is the thing I will be looking at and interacting with.

One European car that I find really ugly to look at is the Porsche Cayenne [2]. Most disagree with me on this. I can't put my finger on why exactly but all of the proportions just looks slightly wrong to me.

EDIT: I've replaced Cayman with Cayenne as pointed out in the comments.

1. http://gmauthority.com/blog/wp-content/gallery/2017-chevrole...

2. https://s1.cdn.autoevolution.com/images/news/gallery/2018-po...

The Cayenne looks ugly because it's such a blatant "tick all boxes" concept - We need an SUV, but we're Porsche, so let's run a halfway morph between a Range Rover and a Carrera and call it a day. Quite like the back half of the Panamera...

*Cayenne - as pointed out by comment above

Thanks, updated now

Dunno about that, the 1st generation Cayenne looks ugly, the later ones look much better. The Panamera is pure ugly all around.

> One European car that I find really ugly to look at is the Porsche Cayman [2].

That's the Cayenne. Cayman is a two-seater.

Thanks, I've updated it now.

I assumed that it was a attempt by someone at Porsche to undermine the long term popularity of SUVs by creating one that is eye-wateringly ugly.

Plenty of yanks hate the plastic whales, too.

I live in the home of the big 3, the metro Detroit area, and there's a pretty healthy mix of all nationalities of cars, so I don't think there's anything profoundly "American" about a Cadillac. In wealthy areas around here, I would even say that tastes run decisively German.

But car design is essentially dying as a significant differentiator. Consumers' preferences are quickly converging on jellybean-like mini SUV's - it's hard to quantify, but cars are more than ever about utility and not identity. To the extent that non-SUV's even matter any more, manufacturing technology improvements are making mid-range Korean cars about as visually striking as an Audi to the average person.

Anyway, I mostly suspect that recognizably "American" cars are holding on to some garish/aggressive elements because they're the the only way left to visibly differentiate - they'd rather be spunky and recognizable as their own (ugly) brand than blend in with Hyundais and Mazdas on the road.

I think what I mean is that even non-American brands will adjust their cars for the American market in such a way that to my eyes, it makes the car super ugly - but I guess it must be making them attractive to Americans, and I would love to understand why/how.

Case in point:

2013 Nissan Qashqai, as seen in EU:

https://www.ledperf.co.uk/images/models/W500/qashqai_nissan_...

2013 Nissan Rogue(so Qashqai adjusted for the American market):

https://cars.usnews.com/static/images/Auto/izmo/353341/2013_...

It's the exact same car. Yet the American version gets this attrocious front, the overgrown grille and these lights that don't really go anywhere and that don't look like anything.

Again, this is my European taste speaking - if any Americans read this and they find the Rogue attractive but the Qashqai horrible - please, please tell me why it is so.

The Rogue is the US "x-trail", not the US Qashqai! It's one size up

(but indeed they seem to have followed the godawful Murano styling on the US Rogue/XTrail)

It is nowadays - 2017 Nissan Rogue is the European X-Trail. It wasn't the case in 2013 - it was just a qashqai. I don't think the previous gen X-Trail was even available in the US at all.

This is particularly visible on the inside:

2013 Rogue:

http://st.motortrend.com/uploads/sites/10/2015/09/2013-Nissa...

2013 Qashqai:

http://3.bp.blogspot.com/-HF9QSCoWifs/UcL9LU3c7FI/AAAAAAAAQE...

I happen to own a 2013 Nissan X-Trail and it's a completely different vehicle, it's a blocky offroader, not an inflated Qashqai like it is nowadays.

Ah I think it's the "2017 Rogue Sport" which is a Qashqai. The Rogue remains an X-Trail (which has indeed been watered down since the 08MY offroad style)

funny that they would follow the same pattern as JLR, releasing a "sport" version which is actually based on a lower category - See 1st gen Range Sport

And then, because obviously it wasn't confusing enough, in Canada the Qashqai is a Qashqai, not Rogue Sport, but X-Trail is still a Rogue.

http://www.nissan.ca/en

To be fair, the compact/subcompact SUV craze we see here is Europe is a bit of a design fad too - most of these have no significant advantage over the equivalent compact hatchback, and cost more to buy and run.

If people wanted utility, they'd either be in hatches or LAVs (dokker, verso, and all the panel-van derivates: kangoo, partner etc)

(I went halfway and got a yeti - awesome vehicle)

I was reading your comment and thinking - "My Yeti has lots of advantages over a hatchback" - then I saw your last line ;-)

Just got a new one, bit gutted that they are phasing them out - the I was so attached to the last one that I was actually quite sad when I traded it in, which was a first for me!

Same here - I can't believe they're replacing it with a bland Ateca clone...

Indeed it's a bit of a compromise between the sensible Roomster and the pretty but useless SUVs

There's plenty of people in the US who agree with you...

That is why German, Japanese (Toyota, Honda, Subaru, etc), and even South Korea (Kia) vehicles are fairly popular in the US too. They have a similar look to vehicles produced for the European market.

I absolutely agree that the traditional US-based manufacturers such as GM, Chrysler, and Ford have a certain ultra-macho-aggressive look to them. There is a certain demographic that likes to buy American manufactured vehicles, and also likes that type of styling.

Tesla is actually strange because they're an American manufacturer that mostly appeals to liberals.

Since Europeans are much more sophisticated with their culture and tastes then us simpleton Americans it makes total sense.

Say did you hear about the European successful Mars rover or maybe any interesting technological innovation coming out of Europe in the last 50 years?

But seriously the CT6 compares nicely to the top range 7 series and S class Mercs. It has a bit of flair but in America we see that as a good thing.

Look, I don't mean to offend anyone. Like I said, I'm genuinely curious if there's any research into why Europeans on average find American cars unattractive, and (if that's the case) why Americans find European cars unattractive. There must be some design elements that resonate with certain groups and I would like to know what they are.

> I find them bloated, full of plastic, tacky, tastless

To be fair, if HN comments are any guide, this describes how most Europeans feel about most aspects of American culture.

> not to mention that it only works on 'vetted' stretches of road.

To be specific GM is following the NHTSA guidelines for autonomous vehicles, which consider Operational Design Domains in the design process.

If I had money to invest, I would put some of it into highway paint producers. All the vision systems depend on crisp high-contrast lane markings, so states will soon be under pressure to re-stripe their highways so their citizens can use their autonomous cruise equipped cars.

You just don't like the way it looks because its not a tesla.

The CT6 is a beautiful car with nice materials and good build quality, you should take a look at it up close.

"If none of that succeeds in returning the driver to the task of controlling the vehicle, it will slow to a stop in the lane of travel, put the hazard flashers on, and summon help via OnStar."

I presume they call OnStar since once the car stops in a live lane on the Interstate, the next thing that will happen will be getting rear-ended by a truck.

That does seem a pretty dangerous action. But I guess, if the vehicle doesn't have any real context/understanding of "get off the road if possible" (after all, even lane changes are manual) and the driver is apparently unresponsive, that's pretty much the only option left.

The car is likely thinking the person has already suffered a medical event, rather than pre-calling for the impending truck.

I think 'fully' automated self driving will only become a reality once 2 things happen:

1) Inter-vehicle communications is standardized. This allows for each vehicle to communicate to each other and signal important metrics (speed,direction,intent...), also allows to complement each vehicles 'knowledge' and context of what is going on on the road at any given moment. Communication only needs to happens for vehicles clustered in a given radius (depending on speed and flow of traffic).

2) A critical mass of autonomous vehicles are on the roads.

Once you have these 2 conditions it will be easier for the 'swarm' behavior and flow of traffic to be automated fully. It will also remove a lot of the 'unknowns' stemming from human drivers and smoothly direct traffic.

Why? Humans drive today without any of this, and with much less sensory perception than 360-degree cameras, radar, lidar, ultrasonics, GPS anchoring, and digital maps. What's the limitation on being able to synthesize these inputs and respond to them?

I don't think machines currently have enough contextual information to be able to drive fully automated. Case in point: the limitations mentioned in the article.

I don't think you can compare human cognition and driving abilities to current AI/Hardware, since the roads and all the information to drive are geared towards humans.

This being said: I do believe fully automated cars will come, but the 2 conditions I mentioned are prerequisites (in my opinion).

Computation power and reliability.

Visual identification algorithms are not fast enough to keep up with driving, let alone distance, direction, and speed inference. Speeding them up is a largely parallel job, but cars don't have the space/power/temperature-controlled/vibration-free environment to support the necessary hardware.

> Inter-vehicle communications is standardized

That is a recipe for easy sabotage. I think any self-driving car can accept inputs from all sorts of senors and/or peer devices, but at the end of the day it should "trust no one".

Everything around you is ripe for sabotage, even today's cars: https://www.wired.com/2016/08/jeep-hackers-return-high-speed...

Inter-vehicle communication is not immune from that also. But it is a necessary condition, in the same way the air traffic control towers help orchestrate flight and access to airspace.

> air traffic control towers help orchestrate flight and access to airspace

those comms are controlled by humans, and they are not hackable. If you want an air traffic controller to transmit dangerous instructions to a pilot, you must kill the good air traffic controller and replace him with a saboteur.

Did you not see Die Hard 2?

Yes, of course Mr. Wonka.

IIRC, some research suggested that ~20% automated traffic already makes a huge dent in congestion.

TL;DR: The innovation here is that you are officially not required to hold the steering wheel. The requirement to keep your hands on the wheel is replaced by a camera that monitors your eyes to ensure that you are paying attention to the road.

capacitive sensors detect the driver’s hand(s) on the wheel, eliminating the need to wiggle it to verify control as with some lane keep assist vehicles.

Only in situations that demand manual control? I agree that the article is not crystal clear about this detail, but this is how I understand it.

An aircraft carrier can't turn on a dime but when it does finally turn around it's a f-ing aircraft carrier and it means business. GM realized it should care about self driving tech, bought some self driving startups, wrote some checks, made more than a token attempt at integrating it with their products. We'll see how it works out.

From how they've described the situation that the system needs to take effect it sounds well thought out. Only time will tell if it makes a habit of driving into partial lane obstructions at highway speed.

>Curve Speed Control

I seriously hope there's a user setting for how aggressive this is because what Caddilac considers a conservative enough setting for its customers is going to be very restricting to a lot of commuters.

> what Caddilac considers a conservative enough setting for its customers is going to be very restricting to a lot of commuters.

If you're letting the computer drive for you in the first place, I doubt the intensity of G-force in a corner is of paramount concern.

I think having it back off in a corner an then having a bunch of people cut in front of you and then very, very slowly regaining speed (because where there was space now there's people who moved in front of you) will annoy people.

Great, more half-baked "self-driving" that will get people killed.

Use the technology where it can actually save lives, like in auto-braking systems in case of accidents, and so on -- not in "autopilot" entertainment/showing-off systems where the driver is supposed to keep an eye on the road at all times anyway.

The argument is that we are trying to remove humans from the equation. Put fully automated systems on the road that have near-perfect reaction times. The only way to get there is by boiling the frog. Technology is scary. It is even more scary when you tell someone that you're entrusting your life to automated tech. What do you do to change that? Marketing!

Start by using singular feature roll out. We've seen lots of these little systems being introduced: sensors for blind spots, auto-braking, adaptive cruise-control, lane-assist, alerts for drifting out of lane, etc. Now they're telling consumers that driving is becoming automated while auto manufacturers have been slowly automating the driving process feature-by-feature for years. They're up front that it isn't unrestricted autonomy. That is, "Hey, this will keep you safe, but just in case, keep your eyes on the road. Don't worry though; these are the same features you've been using for years. "

This isn't a dumb entertainment system for showing off to your friends. It's field testing for future iterations when it is 100% automated. It's genius.

The self-driving tech. only has to be better than humans to save lives, year over year.

And it's been introduced in auto-braking, lane keeping, and other tech. slowly. It's all part of a road map toward fully autonomous.

The driver "keeping their eye on the road at all times" is just a legal requirement, since the laws are a big grey area for manufacturers, at the moment.

>laws are a big grey area for manufacturers

I'm not sure they're really all that grey. They're assistive driving systems and it's the responsibility of the driver to be in full control at all times. Now, as these systems get better and more comprehensive, it's reasonable to ask whether that's a realistic expectation. But it is the law.

finally - 1000s fewer drunken fleshbots behind the wheel.

(I am concerned about security/hacking)

Note that Cruise Automation did not contribute in any meaningful way to GM Super Cruise.

GM's engineers are shipping a real product.

At Cruise Automation we're writing software that is so thoroughly unreliable and wrongly designed that it will never be the basis of a real product.

Reply on news.ycombinator.com