BMW made a self-driving motorcycle • Hacker News Board

Re: BMW made a self-driving motorcycle

#21

Post by craftyguy » Thu, Sep 13, 2018, 3:12 PM UTC

hprotagonist wrote: post #16

icc97 wrote: post #9

Earlier quoted context omitted.

Crash is a better term than accident [0]. > the “one party is going straight and the other is turning left at a four way” accident A self-driving motorcycle would solve that problem, because where as a motorcyclists vision is significantly impaired from the helmet, detection systems on the bike can be looking in all directions [0]: https://www.citylab.com/transportation/2015/09/why-we-say-ca...

I agree that crash is the proper term, but i just forgot. >, because where as a motorcyclists vision is significantly impaired from the helmet... This is incorrect. My vision on the bike is not impaired by my helmet. My field of view in a car is more restricted -- and critically in SMIDSY situations, the peripheral field is obstructed by the pillars holding the roof up. SMIDSYs happen when the car fails to give way t…

> but we solve them by designing intersections such that they can't happen.

How? I'm a cyclist who also has to contend with these same problems..

Re: BMW made a self-driving motorcycle

#22

Post by konschubert » Thu, Sep 13, 2018, 3:31 PM UTC

icc97 wrote: post #11

The point of this motorcycle is to take over if it spots a potential crash scenario. This kind of passive system is the perfect kind of system for cars as well as motorcycles. Self-driving systems are great for alertness/response time, where as humans are better at vision. So having the self-driving systems take over if the human misses something / becomes unwell is the perfect kind of human augmentation. This is opp…

This is a very insightful comment.

Re: BMW made a self-driving motorcycle

#23

Post by jamestimmins » Thu, Sep 13, 2018, 4:36 PM UTC

I would have assumed most motorcycle accidents come from other drivers not seeing the motorcycles. In which case it seems like a helmet with multidirectional cameras and a warning system (eg alerts if you're in another car's blind-spot, or if the car in front of you has stopped rapidly) would have more impact than modifying the motorcycle.

Re: BMW made a self-driving motorcycle

#24

Post by opless » Fri, Sep 14, 2018, 2:33 AM UTC

icc97 wrote: post #19

hprotagonist wrote: post #16

Earlier quoted context omitted.

I agree that crash is the proper term, but i just forgot. >, because where as a motorcyclists vision is significantly impaired from the helmet... This is incorrect. My vision on the bike is not impaired by my helmet. My field of view in a car is more restricted -- and critically in SMIDSY situations, the peripheral field is obstructed by the pillars holding the roof up. SMIDSYs happen when the car fails to give way t…

> My vision on the bike is not impaired by my helmet. I've never worn a motorcycle helmet, but the visor seems so small that you'd lose some of the peripheral vision. I agree that it would often be the car at fault, but as a cyclist I wouldn't care who was at fault I'd just want to avoid the crash. I'd only crash into the car if my peripheral vision missed the car. > we solve them by designing intersections such that…

Having been a motorcyclist when I was younger and ostensibly more reckless (not that motorcyclists are more reckless) I can tell you from experiance that you have very little getting in your way visually. However in a car you have two pillars holding your roof up to your immediate left and right, and door frames and boot pillars over either shoulders.

As a motorcyclist you have way more situational awareness than other vehicles on the road.

You also quickly learn that other road users are blissfully unaware of you and you drive accordingly.

Re: BMW made a self-driving motorcycle

#25

Post by hprotagonist » Sat, Sep 15, 2018, 8:10 AM UTC

craftyguy wrote: post #21

hprotagonist wrote: post #16

Earlier quoted context omitted.

I agree that crash is the proper term, but i just forgot. >, because where as a motorcyclists vision is significantly impaired from the helmet... This is incorrect. My vision on the bike is not impaired by my helmet. My field of view in a car is more restricted -- and critically in SMIDSY situations, the peripheral field is obstructed by the pillars holding the roof up. SMIDSYs happen when the car fails to give way t…

> but we solve them by designing intersections such that they can't happen. How? I'm a cyclist who also has to contend with these same problems..

for cyclists (and i’m a cyclist too), the dutch basically have this solved. You do this: https://bicycledutch.wordpress.com/2018/02/20/a-common-urban...

For motorcycles, you mostly only have to ensure turn arrows as well as normal stoplights. Permitting lane-splitting also helps.