Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

George Hotz cancels his Tesla Autopilot-like ‘comma one’

electrek.co

Post reply on HN
Print view View original

261–270 of 498 posts

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#261

Post by module0000 »

devy wrote: post #110
hangonhn wrote: post #25

Wow. If he really moves his project to China to take advantage of a lax regulatory environment even though he couldn't pass the scrutiny in the US, he's being very immoral and cavalier about people's lives in China. I don't understand how a startup like Cruise can get past this regulatory hurlde in the US but Hotz can't.

I am no expert in Chinese transportation laws and regulations. But I wouldn't say it's a lax environment there - it's more ad hoc to be more accurate as I've seen regulation coming out from time to time in response to events. On the other hand, Shenzhen is the world capital of electronic / hardware components sourcing, manufacturing & production. Geohot's trip to Shenzhen may or maynot be an indication of his intenti…

If China's transportation regulations are anything like their import/export regulations..."lax" isn't the right way to describe them, more aptly described as "influenced in your favor by money paid to the right people".

Disclaimer: Worked for a company exporting an "organic" product from China - which meant paying the harbor master a rather large fee per ship to have them labeled as such. When the product arrives in the US and other countries, it's treated as being whatever it is labeled as.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#262

Post by DannyBee »

logfromblammo wrote: post #178
tptacek wrote: post #171

Earlier quoted context omitted.

"Kissing the ring"? They want his test documents, not his fealty.

They want his compliance, or his $21k/day. The document is all stick and no carrot. And it is a big, federal regulatory agency stick. I think "kissing the ring" applies.

You also they often have very common procedures for these types of letters, right? Usually brought on by court cases, etc. For example:

http://www.fda.gov/ICECI/ComplianceManuals/RegulatoryProcedu...

They also generally warn of fines ahead of time, because otherwise they get people arguing due process.

Seriously, this is not that uncommon. Note the fine warnings in all of these special orders:

http://www.documentcloud.org/documents/1349845-n-h-t-s-a-spe...

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&c...

http://www.safercar.gov/staticfiles/safercar/recalls/Special...

http://www.autosafety.org/sites/default/files/imce_staff_upl...

etc

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#263

Post by tptacek »

patio11 wrote: post #257
ceterum_censeo wrote: post #246

Earlier quoted context omitted.

The documentation that the NHTSA is requesting are things that would already exist for a properly designed product. They are not asking for it to be generated de novo (and in fact, it cannot be done in that manner if they are operating under a proper design process). This letter is designed to give pause to (e: previously I wrote "kill") a company operating with complete disregard for proper operations in a regulated…

This letter is designed to kill a company operating with complete disregard for proper operations in a regulated industry. We are in entire agreement, but people cannot believe in the superposition of this and "Oh that's just some simple questions from the regulator." Edit to add: Quoted bit appears to have changed after my quote. I think the old version remains accurate.

They're asking for documentation that any serious vendor would already have. That we believe this would kill comma.ai is really more a testament to our priors that George doesn't keep anything resembling rigorous test documentation.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#264

Post by argonaut »

shraken wrote: post #138

I'm not involved in the machine vision field but I found his commitment to open and transparent academic publishing very admirable. He published a paper with their summer intern here: https://arxiv.org/pdf/1608.01230.pdf

Also, the entire deep learning academia is already extremely open and transparent with publishing.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#265

Post by ryanwaggoner »

ryanmarsh wrote: post #259
abduhl wrote: post #144

Earlier quoted context omitted.

"As you are undoubtedly aware, there is a high likelihood that some drivers will use your product in a manner that exceeds its intended purpose." The government appears well aware of this hand waving garbage that companies are trying with "it isn't really autopilot, keep your hands on the wheel wink ." This is exactly government's job and is a fabulous example of a government reacting to ongoing events that have prov…

Yes. There are a number of federal agencies that are a shit-show at best but I can't remember a time in my adult life when I was disappointed in NHTSA.

The NTSB also does amazing work, though it's the smallest Federal agency (~400 people), which I think probably helps greatly.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#266

Post by Matthias247 »

jacquesm wrote: post #72
tptacek wrote: post #29

Here's what the Special Order demanded: 1. Describe in detail how the comma one is installed in a vehicle and provide a copy of installation instructions for the comma one. 2. Describe in detail the advanced driver assistance features of the comma one, including how those features differ from the existing features of the vehicles in which the comma one is intended to be installed. 3. Describe in detail how a vehicle…

That's an incredibly reasonable list and something anybody active in the self driving car space should be more than happy to answer in the required detail.

I agree. Imho it even sounds more like a questionaire that provides them an overview of the system rather than a really extensive safety analysis.

Even the simplest (non-safety-critical) systems in automotive requires typically a lot more documentation. E.g. if you have to prove that your development process is automotive spice compliant then you need documentation for everything - from detailled requirements to test conformance. For safety critical stuff it's even more.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#267

Post by knodi123 »

jacquesm wrote: post #59
Analemma_ wrote: post #16

He's currently throwing a temper tantrum on Twitter, and all because the NHTSA wanted to ask him a few questions about the product (you can see it here: https://www.scribd.com/document/329218929/2016-10-27-Special... ). Like, how dare the government try to have basic standards for multi-ton hunks of metal hurtling down the highway without human input. I've been trying to convince people lately that Silicon Valley isn…

I tweeted a joke yesterday that the Engineering Society had changed it's slogan to 'Move fast and kill people', I'm pretty happy that NHTSA has its feet solidly planted on terra firma and requires a bit of process around releasing something as important as self driving vehicles. The more software becomes mission critical the more we will need level headedness and quality rather than tantrums.

> I'm pretty happy that NHTSA has its feet solidly planted on terra firm

Makes sense, since they regulate the activity that is the 4th largest cause of death in the US.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#268

Post by 8_hours_ago »

Do third-party "autopilot" systems require ISO 26262 certification? I know that car manufacturers and tier-1s spend lots of time and money getting their safety-critical code certified. Do the same requirements not apply to aftermarket solutions?

Actually, now that I think about this more... does Tesla's autopilot have ISO 26262 certification? I hope that the recent advancements in self-driving cars isn't a result of tech companies bending or breaking the rules, but I don't know enough about when ISO 26262 is required (if at all).

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#269

Post by dmritard96 »

erikpukinskis wrote: post #180
abduhl wrote: post #144

Earlier quoted context omitted.

"As you are undoubtedly aware, there is a high likelihood that some drivers will use your product in a manner that exceeds its intended purpose." The government appears well aware of this hand waving garbage that companies are trying with "it isn't really autopilot, keep your hands on the wheel wink ." This is exactly government's job and is a fabulous example of a government reacting to ongoing events that have prov…

Proof that so-called autopilot systems are a danger to the public would be evidence that they perform worse than the average driver. I haven't seen such evidence, what are you referencing?

beyond the obviousness of what you have overlooked, its also worth pointing out that not all autonomous car usage will be substitutive. What if i just send my autonomous car to go take a picture of something? I might not have done it before, but now I will because it only costs a few cents/mile and none of my time.

Re: George Hotz cancels his Tesla Autopilot-like ‘comma one’

#270

Post by claudius »

assface wrote: post #209
shraken wrote: post #138

I'm not involved in the machine vision field but I found his commitment to open and transparent academic publishing very admirable. He published a paper with their summer intern here: https://arxiv.org/pdf/1608.01230.pdf

This is not peer-reviewed. This is a tech report. Anybody can post these.

Well, you usually need a university address or some endorsers to post to the arxiv, but true, most people could probably post that.

On the other hand, the arxiv is also the default method of communication in quite a few fields (physics and maths, mostly), to the point that while grant committees etc. look at your peer-reviewed papers, I essentially only check the arxiv for new developments, not the plethora of properly peer-reviewed journals.

Post reply on HN