[SystemSafety] [External] NTSB report on Über fatal accident in Tempe

[Peter Bishop]

I also have difficulty with the use of classification.
It is really hard to get it right all the time
- and getting it wrong could be fatal.

For maximum safety, what we need is something simple and reliable
(i.e. no fancy vision processing or machine learning), e.g.

Can we see a warm blob? (infra red)
     How far away? (radar)
     If close enough (wrt to braking distance)
     - Is it moving? (IR+radar)
     - continuously extrapolate future position of object based on its
current speed (IR+radar)
     - react if collision possible

For cold objects (IR+radar),
    - ignore static objects not on planned path
     If close enough (wrt to braking distance)
     - continuously extrapolate future position of object based on its
current speed and direction
     - react if collision possible

I suppose you could use lidar rather than radar, but radar is probably
less complicated and less affected by ambient conditions (fog, rain,
etc.). It would need to be a scanning radar though to get both distance
and direction. Lidar is more accurate - but does it matter if you want a
miss rather than a hit?

Peter Bishop

On 10/11/2019 17:50, Ug Free wrote:
> The classification is mandatory to at least  predict correctly the next move of the objects: without moves and dynamics, the problems are much easier! 
>
> best regards
> Hugues
>
>> On 10 Nov 2019, at 12:31, Grazebrook, Alvery AN <alvery.grazebrook at airbus.com> wrote:
>>
>> 
>> -----Original Message-----
>> ...On Behalf Of Driscoll, Kevin R
>>
>> I never have understood the autonomy community’s preoccupation with classification of potential collision objects.  Am I being naïve, or there are just (currently) two classes of objects:  those posing a collision hazard and those where “collision” is OK (generally things that have low enough mass that they can be windblown, such as snow* or plastic bags).  Eventually, autonomous systems will become smart enough to be able to tackle Trolley Car Dilemma decisions requiring sophisticated classification, but I think that’s a long, long, long way off.
>> [Grazebrook, Alvery AN] 
>> To be honest, I'd be happy that an autonomous car was able to clearly identify cases such as windblown / drifting snow, and refuse to drive autonomously in those conditions (or any other) where the autonomous function was likely to struggle. I appreciate there are parts of the world where this is going to undermine their market acceptance, and there's a chance that the vehicle will strand someone by being unable to complete a journey. 
>>
>> It slightly bothers me that the debate around these vehicles doesn't say much about the "we aren't going there" part of the story. Clearly this is a limit that will vary as the systems mature. I guess implementing this will depend heavily on classification schemes. 
>>
>> Cheers,
>>    Alvery
>>
>> ** These opinions are my own, not necessarily those of my employer
>> This email and its attachments may contain confidential and/or privileged information.  If you have received them in error you must not use, copy or disclose their content to any person.  Please notify the sender immediately and then delete this email from your system.  This e-mail has been scanned for viruses, but it is the responsibility of the recipient to conduct their own security measures. Airbus Operations Limited is not liable for any loss or damage arising from the receipt or use of this e-mail.
>>
>> Airbus Operations Limited, a company registered in England and Wales, registration number, 3468788.  Registered office:  Pegasus House, Aerospace Avenue, Filton, Bristol, BS34 7PA, UK.
>> _______________________________________________
>> The System Safety Mailing List
>> systemsafety at TechFak.Uni-Bielefeld.DE
>> Manage your subscription: https://lists.techfak.uni-bielefeld.de/mailman/listinfo/systemsafety
> _______________________________________________
> The System Safety Mailing List
> systemsafety at TechFak.Uni-Bielefeld.DE
> Manage your subscription: https://lists.techfak.uni-bielefeld.de/mailman/listinfo/systemsafety

-- 

Peter Bishop
Chief Scientist
Adelard LLP
24 Waterside, 44-48 Wharf Road, London N1 7UX

Email: pgb at adelard.com
Tel:  +44-(0)20-7832 5850

Registered office: 5th Floor, Ashford Commercial Quarter, 1 Dover Place, Ashford, Kent TN23 1FB
Registered in England & Wales no. OC 304551. VAT no. 454 489808

This e-mail, and any attachments, is confidential and for the use of
the addressee only. If you are not the intended recipient, please
telephone 020 7832 5850. We do not accept legal responsibility for
this e-mail or any viruses.