revtothelimit - motorbike forum

The Spin Doctor wrote: Fri Sep 23, 2022 3:48 pm

slowsider wrote: Fri Sep 23, 2022 2:19 pm

The Spin Doctor wrote: Fri Sep 23, 2022 1:42 pm How many include motorcycles doing mad stuff like the flat-out Deliveroo rider who swerved out from the blindspot behind a van just as I was pulling out of a side turning 10 mins ago?

I have no idea why he was riding in the gutter there as the road was clear!

Did you get a surprise?

As a driver, it was closer than I would have liked. From his perspective? Not really even a near miss, since he only had to roll off a tad. Certainly, nothing much to have bothered me when I had my courier head on!

To answer your first Q, I've no idea However, if it's open access and searchable, then you might be able to check

If not, suggest it

For your second post, you raise an interesting challenge: how are situations assessed for risk? If, with courier head on, you were assessing test situations, would you have a much smaller database? From that, how should safety of AVs be assessed, by whose 'standards'?

"Needs more work"

Here's a question I heard (in the context of AVs):

How safe is 'safe'?

i.e. How does the decision get made (justified, assessed, etc.)?

Is it compared to someone at the stage of passing the Theory, HP and practical driving tests?

If 'no, higher standard, better, safer', how would that be assessed - and how is it justifiable?

So if that 'higher, better', perhaps equivalent to an 'average' driver, what is that level based on?

Horse wrote: Thu Oct 06, 2022 3:50 pm Here's a question I heard (in the context of AVs):

How safe is 'safe'?

i.e. How does the decision get made (justified, assessed, etc.)?

Is it compared to someone at the stage of passing the Theory, HP and practical driving tests?

If 'no, higher standard, better, safer', how would that be assessed - and how is it justifiable?

So if that 'higher, better', perhaps equivalent to an 'average' driver, what is that level based on?

Good question.

slowsider wrote: Thu Oct 06, 2022 4:26 pm

Horse wrote: Thu Oct 06, 2022 3:50 pm Here's a question I heard (in the context of AVs):

How safe is 'safe'?

i.e. How does the decision get made (justified, assessed, etc.)?

Is it compared to someone at the stage of passing the Theory, HP and practical driving tests?

If 'no, higher standard, better, safer', how would that be assessed - and how is it justifiable?

So if that 'higher, better', perhaps equivalent to an 'average' driver, what is that level based on?

Good question.

I said it was a good question, I didn't say it was something that I wanted to pay £40 or £60 to read someone else's answer.

On here's free. What's your answer?

I've only been in an automated vehicle once and, fine as it was for what it was doing at the time, I wouldn't have wanted to be out on a public road in it. But that would have been my decision, then. For reasons I'm not going into.

How would you make your decision?

I've driven cars without a seat belt, and ridden without a helmet but that was my decision. I would have based that on my knowledge and perception of the risks involved. My trips in AVs have been in very limited ODDs.
I have no way of knowing what the algorithms in road-going AVs will decide to prioritise when making choices on the road. Koopman is involved in the standards definitions for safety levels. His opinion might be of more relevance than some random guy on the interweb. Your view on your areas of expertise would perhaps be worth paying for.
You pays your money (or not) and takes your choice.
I'm waiting for a second hand copy, if you want to punt it on after

Some Free stuff
https://users.ece.cmu.edu/~koopman/SafeEnough/

slowsider wrote: Thu Oct 06, 2022 7:33 pm
His opinion might be of more relevance than some random guy on the interweb.

His opinion is valuable for engineers, etc.

But for the general public (like those cab clients), it is the randoms' views that are important.

I've been on one or two dodgy rides with professional drivers ...

slowsider wrote: Thu Oct 06, 2022 7:33 pm
Your view on your areas of expertise would perhaps be worth paying for.

If I could remember my charge-out rate

Fail

https://www.bbc.co.uk/news/uk-scotland- ... s-63158295

That's reassuring.

Mr. Dazzle wrote: Sat Nov 05, 2022 5:12 pm

https://www.npas.police.uk/news/man-ple ... helicopter

Slightly lower-tech, tho'

I've had a laser pen pointed at me a couple of times... very dazzling when it gets you in the eyes.

Mind you, the cat loved chasing mine

A lot might depend on whether the AV had already identified and reacted to the ped. There's work going to go beyond identification and on to understanding and prediction.

And, just a guess, if lidar systems are used in any military / security application, then it's probably a known vulnerability and someone will have a fix.

Dare I mentioned 'added cost and complexity'?

I appreciate that technology CAN make things safer, but it appears in this case that the vehicles are using untried tech, and essentially testing them live. .

It appears Telsa are attempting to dodge scrutiny... this story is actually from the summer but it's the first time I've seen it. Basically, investigators have forced Tesla to disclose more data about the crashes their vehicles have been involved in, not just ones declared as happening when autopilot was running... and examining the data...

"Federal investigators step up probe into Tesla Autopilot crashes

"The new data set stems from a federal order last summer requiring automakers to report crashes involving driver assistance to assess whether the technology presented safety risks. Tesla‘s vehicles have been found to shut off the advanced driver-assistance system, Autopilot, around one second before impact, according to the regulators.

"The NHTSA order required manufacturers to disclose crashes where the software was in use within 30 seconds of the crash, in part to mitigate the concern that manufacturers would hide crashes by claiming the software wasn’t in use at the time of the impact."

“It revealed that more crashes are happening than NHTSA had previously known,” said Phil Koopman, an engineering professor at Carnegie Mellon University who focuses on autonomous vehicle safety. He noted that the reports may omit more minor crashes, including fender benders.

https://www.washingtonpost.com/technolo ... t-crashes/

https://www.cnn.com/2021/08/16/business ... index.html

Tue August 17, 2021

Federal safety regulators are investigating at least 11 accidents involving Tesla cars using Autopilot or other self-driving features that crashed into emergency vehicles when coming upon the scene of an earlier crash.

https://amp.cnn.com/cnn/2022/06/09/cars ... index.html

Thu June 09, 2022

Federal investigators are expanding their probe of Teslas that have slammed into parked first responders' vehicles.

The National Highway Traffic Safety Administration said Thursday its investigation now encompasses 830,000 vehicles with similar driver assist technology. Investigators have reviewed approximately 200 Tesla crashes and are focusing on 16 crashes involving emergency or road work vehicles.

I honestly haven't read all of the thread, but, how many of the incidents whilst using autopilot or self-drive stuff are due to drivers being stupid? Like the advert -

why are you leaving the driver seat to go make a cup of tea?
Well, the camper has self-drive, so I don't need to be there!

Is it even possible to to regulate against that sort of stupid?

Noggin wrote: Tue Nov 15, 2022 8:36 am
Is it even possible to to regulate against that sort of stupid?

Not without a culling