General Automation Thread

N&P: Discuss governments, nations, politics and recent related news here.

Moderators: Alyrium Denryle, SCRawl, Thanas, Edi, K. A. Pital

Post Reply
User avatar
madd0ct0r
Sith Acolyte
Posts: 5597
Joined: 2008-03-14 07:47am

Re: General Automation Thread

Post by madd0ct0r » 2018-12-02 03:34pm

Im thinking about this s lot at work.

Theres a real trap ahead. We need to automate carefully. The current model for cars hits this - the stage where humans do almost nothing but need to react carefully in emergencies simply does not work.

We need to automate in a way to avoid these sort of local minima.
"Aid, trade, green technology and peace." - Hans Rosling.
"Welcome to SDN, where we can't see the forest because walking into trees repeatedly feels good, bro." - Mr Coffee

User avatar
FaxModem1
Emperor's Hand
Posts: 6762
Joined: 2002-10-30 06:40pm
Location: In a dark reflection of a better world

Re: General Automation Thread

Post by FaxModem1 » 2018-12-02 04:08pm

Elheru Aran wrote:
2018-11-30 12:44pm
FaxModem1 wrote:
2018-11-29 09:47pm
Elheru Aran wrote:
2018-11-29 06:03pm
I'm guessing said warehouse is still going to have human employees, for goods that robots may not be able to handle easily such as fresh produce... which in retrospect probably doesn't get warehoused much, it's probably shipped relatively direct with few links in the chain from farm to store in order to keep it fresh. So... yeah, I dunno. I do think there will still be humans working there in tasks that they don't have robots for yet, unloading trucks for example. Of course, they could just palletize all the stuff on the truck and thus a robot forklift driver would just pull it all out like that... I don't know.

I'm a bit bearish on the whole robot thing, at least in the immediate future (by which I mean roughly the next five-ish years). Self-service kiosks and such don't really qualify as 'robots' to me but they might apply, I suppose-- those I do see proliferating, particularly with the increasing technological interconnectedness of First World services.
When you say you are bearish, do you mean that you doubt companies will be investing in cheaper machines such as robots, or that you don't think that it's the right terminology?

Because in regards to terminology, we're hitting the point where I think what's a program, what's a machine, and what's a robot is starting to get a bit mixed in regards to production models.
Bit of both.

As far as terminology goes, while selfservice kiosks and such do fall under 'automation' I suppose, they aren't 'robots' to me in that they don't particularly... DO anything. Some of them come with a conveyor belt now that you can load your groceries on as you check yourself out, but that's about it. A 'robot' to me has to actually do something outside of hands-on human direction, whether it's moving frozen burgers on a conveyor belt through a heater or stacking boxes of biscuit mix in a warehouse. But that's just me being nitpicky about the term 'robot'.

Regarding the rise of automation in general... I think it's going to depend heavily on the company and the market. Japan for example, they could easily have a whole bunch of mass market commercial robots pretty soon. A backwoods corner of rural USA? Nah. It's going to depend on where the profits are, and on what the customers are comfortable with. Would women, for example, be okay with a robot designed to measure their bust size for bras in the Victoria's Secret? Or would they prefer to have an actual hew-mon person slipping the tape under their arms? Versus the 'burger flipper', which can be mostly squirreled away in the back of the house and customers don't have to care.

But in general I don't know how *quickly* automation is going to be embraced at the retail level. Certainly there will be markets and areas where it pops up and proliferates quickly, tech towns like Seattle and San Francisco probably being good examples, but outside of those areas, it'll depend on profit margin.
Well yes, some areas are going to look rather cyberpunk in a decade or so while a mostly rural community is going to look like....a mostly rural community.

But I can see a global effect economically, and wonder if it will hit that rural community sideways in other ways. And what those ways will be.
madd0ct0r wrote:
2018-12-02 03:34pm
Im thinking about this a lot at work.

There's a real trap ahead. We need to automate carefully. The current model for cars hits this - the stage where humans do almost nothing but need to react carefully in emergencies simply does not work.

We need to automate in a way to avoid these sort of local minima.
What would be the best way to do this without limiting progress? Are there areas in which we should say, humans don't need to work here, it's better for machines to do all the heavy lifting completely, and only have a few overseers to make sure things are fixed and everything is running smoothly?

And how heavily employed are those areas?
Image

User avatar
FaxModem1
Emperor's Hand
Posts: 6762
Joined: 2002-10-30 06:40pm
Location: In a dark reflection of a better world

Re: General Automation Thread

Post by FaxModem1 » 2018-12-08 05:21pm

On the self-driving car front:

Ars Technica
Even self-driving leader Waymo is struggling to reach full autonomy
After 48 hours we haven't seen any sign people are using Waymo's service.
TIMOTHY B. LEE - 12/7/2018, 10:55 AM

Even self-driving leader Waymo is struggling to reach full autonomy
Enlarge
Waymo
115
WITH 85 POSTERS PARTICIPATING, INCLUDING STORY AUTHOR
SHARE ON FACEBOOK
SHARE ON TWITTER
The Wednesday rollout of Waymo One, Waymo's commercial self-driving taxi service, falls far short of expectations the company itself set earlier in the year.

In late September, a Waymo spokeswoman told Ars by email that the Phoenix service would be fully driverless and open to members of the public—claims I reported in this article.

We now know that Waymo One won't be fully driverless; there will be a driver in the driver's seat. And Waymo One is open to the public in only the narrowest, most technical sense: initially it will only be available to early riders—the same people who have been participating in Waymo's test program for months.

This seems to be the latest sign that Waymo's technology is progressing more slowly than a lot of people expected—including Waymo's own leadership a year ago. People who have observed Waymo's vehicles on public roads in recent months report that the cars still struggle with unprotected left turns, merges, and other tricky situations.

FURTHER READING
Waymo’s ambitious plans for high-speed taxis could be holding it back
Waymo is widely seen as the industry leader. The company began working on self-driving technology in 2009, long before most other technology and car companies started taking it seriously. So if Waymo isn't ready to launch a fully driverless service after more than 18 months of intensive public testing, that should make us skeptical of claims from other companies that they'll be ready to launch fully self-driving technology any time soon.

Waymo One is barely a public service
The launch of Waymo One feels less like the launch of a public, commercial service than a rebranding of its testing program. Waymo vowed to launch a commercial service before the end of the year, and Waymo One technically qualifies. But the service hardly seems more open to the public than the early rider program Waymo had last week.

In a Thursday phone conversation, a Waymo spokeswoman declined to tell me how many customers had signed up for Waymo One on the first day or how many rides they'd taken. If a lot of people were signing up, you'd expect some of them to post photos or videos on social media, but 48 hours after the official launch I haven't been able to find any sign of people using Waymo One—and neither have other people.

I'm sure Waymo is telling the truth when it says it has invited "hundreds" of early riders to switch to Waymo One, and I assume at least a handful of people have signed up. But this is not what normally happens when a big public company launches a major new product.

Waymo abandoned plans for a fully driverless launch
This is a screenshot from Waymo's <a href="https://www.youtube.com/watch?v=aaOB-ErYq6Y">November 2017 video</a> announcing the start of fully driverless testing. It shows fully driverless Waymo cars driving on residential streets that are almost empty.
Enlarge / This is a screenshot from Waymo's November 2017 video announcing the start of fully driverless testing. It shows fully driverless Waymo cars driving on residential streets that are almost empty.
Waymo
I wasn't the only reporter who was told that Waymo's first commercial service would be fully driverless.

In March, The New York Times reported that Waymo's Phoenix service would launch by the end of the year. Waymo "intended to move forward rapidly with a driverless ride service for the public because it was confident its vehicles could operate safely in virtually any driving situation where they would be put into use," the Times' Neal Boudette wrote. The same article stated that Waymo expected to be "providing as many as one million rides a day" by the end of 2019. (Update: a Waymo spokeswoman tells us that it never told the Times this.)

FURTHER READING
Waymo One, the groundbreaking self-driving taxi service, explained
But in recent months, Waymo seems to have lost confidence in its fully driverless operations. The Information's Amir Efrati reported last week that "due to concerns about safety, the Alphabet company put so-called 'safety drivers' back behind the wheel of its most advanced prototypes."

Indeed, it's not clear if fully driverless vehicles have ever accounted for a significant fraction of Waymo's testing activities. According to an August story by Efrati, fully driverless tests have typically been "in relatively small residential areas of Chandler, Arizona, where there is little traffic."

That squares with the experience of Ryan Randazzo, a reporter at the Arizona Republic who has covered Waymo over the last two years. Because he's based in the Phoenix area, he's able to regularly observe Waymo vehicles in action and talk to others in the area about the cars—including three days in October and November when he followed around Waymo vehicles to see how they perform.

"I don't know that anyone has ever seen one without a driver in the driver seat" since Waymo first announced driverless testing in November 2017, he told Ars. "They may have done it, but they don't tell us. I've asked five different ways—what percent of rides [are driverless]—they won't answer that question."

Waymo cars struggle with tricky situations
Over the course of October and November, Randazzo spent three days observing Waymo's cars in action—either by following them on the roads or staking out the company's depot in Chandler. He posted his findings in a YouTube video. The findings suggest that Waymo's vehicles aren't yet ready for fully autonomous operation.

"Lane changes appear to be a problem for the cars," Randazzo says in the video. When trying to move into a crowded lane, a Waymo car seemed to lack a human driver's ability to anticipate other drivers' actions and squeeze into an open spot. Instead, the vehicle would turn on its turn signal and wait for a few seconds for an opening to appear. If one didn't appear, it would turn the turn signal off and wait for a while before trying again.

"It actually took this car almost a minute and a half to change lanes here," Randazzo said. "We talked to one person who uses these vehicles around the East Valley. He said that they sometimes miss their turns they're so hesitant."

In another incident, a Waymo car was part of a line of cars approaching an intersection where a car crash was blocking the right lane. The human drivers saw the issue far ahead and began shifting to the left lane. The Waymo car continued straight and only began trying to merge left when it was a few car lengths away from the traffic cones. The car then abruptly cancelled the left merge and instead went right into a turn lane—Randazzo speculates that this was the human safety driver taking over the vehicle.

That jives with a video posted to YouTube in September showing a Waymo vehicle struggling to merge onto a California freeway.

According to The Information's Efrati, Waymo has struggled with turning from a low-speed residential street onto a major boulevard. This is also a problem The Washington Post observed during a recent Waymo test ride. The Post reported that "left turns can be painfully slow" when turning onto a major traffic artery.

With these kinds of challenges, it's not surprising that Waymo chose to keep safety drivers in its cars. And to be clear, I don't fault Waymo for doing this. Quite the contrary, they deserve credit for putting safety first. But if Waymo has largely halted testing with fully driverless vehicles—and there's a lot of circumstantial evidence that they have—it would be nice if Waymo would level with the public about this.

And if even Waymo—long regarded as the industry leader—is struggling to roll out fully driverless cars, that may be a cause for skepticism that its rivals will achieve the feat any time soon. We're going to see more and more self-driving cars on our roads. But it might be a while before we see ones that are fully driverless.
Some delays, but this is a reality.
Image

User avatar
Crazedwraith
Emperor's Hand
Posts: 10568
Joined: 2003-04-10 03:45pm
Location: Cheshire, England

Re: General Automation Thread

Post by Crazedwraith » 2018-12-08 05:32pm

That seems to be the opposite conclusion that the article draws. That the tech is no where near ready.
To the brave passengers and crew of the Kobayashi Maru... sucks to be you - Peter David

User avatar
FaxModem1
Emperor's Hand
Posts: 6762
Joined: 2002-10-30 06:40pm
Location: In a dark reflection of a better world

Re: General Automation Thread

Post by FaxModem1 » 2018-12-08 05:43pm

Crazedwraith wrote:
2018-12-08 05:32pm
That seems to be the opposite conclusion that the article draws. That the tech is no where near ready.
Yes, but it's still amazing to me that we're in a place where they are testing self driving cars. This used to be science fiction, now it's something they're trying to work out the (multitude of) kinks.
Image

Post Reply