[SystemSafety] NHTSA on the Tesla crash

[Nick Tudor]

I was wondering if is any data on the number of collisions/accidents that
didn't happen because of human intervention?


On Mon, 23 Jan 2017 at 06:47, Peter Bernard Ladkin <ladkin at causalis.com>
wrote:

> On 2017-01-20 13:47 , Martyn Thomas wrote:
>
> > All true, but if you sell it as an autopilot ...
>
>
>
> It seems to me that there is already a chance to have things straighter
> for customers. NHTSA has
>
> autodriving function classifications. There is a function called Automatic
> Emergency Braking, AEB.
>
> "Automatic Emergency Braking includes the following crash avoidance
> technologies: Forward Collision
>
> Warning (FCW), Dynamic Brake Support (DBS), and Crash Imminent Braking
> (CIB)" (Section 2, p1, of the
>
> Tesla-crash report).
>
>
>
> From 2007 to 2011 there was a NHTSA-sponsored project undertaken by the
> "Crash Avoidance Metrics
>
> Partnership" to evaluate CIB. "The final report from this project,
> released in September 2011,
>
> validated the effectiveness of radar, camera and radar/camera fusion
> systems as rear-end collision
>
> mitigation or avoidance technologies. The report also identified several
> crash modes that were not
>
> validated by the project, including straight crossing path (SCP) and left
> turn across path (LTAP)
>
> collisions."
>
>
>
> The Tesla crash was an SCP scenario, as far as I understand it. "Not
> validated" in 2011. "Validated"
>
> in 2016? Who knows? There is an opportunity. There is a partnership, a
> classification of scenarios,
>
> and a historical attempt at evaluation. Why not introduce that vocabulary
> and start on standardised
>
> tests with a requirement to advertise your autodriving functions in terms
> of test results? It won't
>
> be perfect. But a manufacturer could say "our AEB scored X/Y on CIB in
> SCP". Purchasers would be
>
> introduced to the difference between avoiding a collision during a merger,
> avoiding an Auffahr
>
> collision, and avoiding an SCP collision.
>
>
>
> Compare with the situation with emissions and fuel efficiency.
> Manufacturers can't just say what
>
> they like in advertising. There is a standard set of tests, and they can
> use those results. The
>
> tests can be and were gamed, as we know. But gaming manifestly has its
> costs. Because of the tests,
>
> the public is regaled with details about NOx and carbon and particles and
> can chat about it over a
>
> beer at the pub.
>
>
>
> One can imagine something similar for autodriving functions. Teenagers
> awaiting their licence and
>
> lusting after "performance" could chat about FCW, DBS and CIB performance
> on SCPs and LTAPs. Just as
>
> today there are different licensing and age requirements for mopeds,
> low-power motorcycles and
>
> general motorcycles, one could imagine they could be different classes of
> licence for four-wheeled
>
> vehicles with different speed and CIB performance. You could get a licence
> at 15 to operate a
>
> vehicle with validated CIB/LTAP and CIB/SCP and ..... Young Joe Smith
> could boast at 17 that he has
>
> just passed his CIB/LTAP and CIB/SCP exemption test.
>
>
>
> PBL
>
>
>
> Prof. Peter Bernard Ladkin, Bielefeld, Germany
>
> MoreInCommon
>
> Je suis Charlie
>
> Tel+msg +49 (0)521 880 7319  www.rvs-bi.de
>
>
>
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
>
> The System Safety Mailing List
>
> systemsafety at TechFak.Uni-Bielefeld.DE
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20170123/db9f5ac6/attachment.html>