self-driving cars and lane splitting [LWN.net]

self-driving cars and lane splitting

Posted Feb 23, 2013 7:37 UTC (Sat) by paulj (subscriber, #341)
In reply to: self-driving cars and lane splitting by giraffedata
Parent article: ELC: Google learns to drive

Filtering by two-wheelers (human or engine powered) is legal in the UK. There is recent case law affirming this.

Also, have to take issue with the the article's claim that the car is amazing, and that before the car it would take "all day" to travel 72 km. Before the car there was the bicycle, and it is within many people's ability, with only a small amount of training, to maintain 15 to 20 km/h on a bicycle on unpaved roads. Which would be just under 5 hours to 3½ hours.

Further, the car's ability to cover distance in reasonable amounts of time is only amazing when relatively few others wish to do so at the same time. When that is not the case, e.g. as is typical for 2 to 5 hours / day in and around major urban areas across many parts of the world, the car's terrible space inefficiency gives an amazing *inability* to make reasonable progress, as everyone sits in pollution spewing traffic jams, with average journey times often lower than they would achieve on a bicycle.

With more and more of us living in urban areas, the personal car becomes ever more impractical.

to post comments

self-driving cars and lane splitting

Posted Feb 23, 2013 14:13 UTC (Sat) by khim (subscriber, #9252) [Link] (10 responses)

Before the car there was the bicycle, and it is within many people's ability, with only a small amount of training, to maintain 15 to 20 km/h on a bicycle on unpaved roads.

Cars actually were invented before bicycle. Centuries before, in fact. Even if you'll narrow-down and exclude all these steam-powered wagons cars still come before bicycle. Compare history of the bicycle with history of the automobile. It's possible to twist definition of bicycle and definition of car enough to somehow make sure bicycle come before cars but it'll be pretty unnatural definitions.

self-driving cars and lane splitting

Posted Feb 23, 2013 15:45 UTC (Sat) by raven667 (subscriber, #5198) [Link]

While an interesting factoid I don't see how it is related to the substance of the previous posters point.

self-driving cars and lane splitting

Posted Feb 23, 2013 18:01 UTC (Sat) by paulj (subscriber, #341) [Link] (8 responses)

Regardless of exactly date very early prototypical automobiles were first made, it's utterly irrelevant to the question of the transport commonly available to people. The bicycle was in mass use *long* before cars were ever produced commercially (and then, in small numbers to begin with, available only to the very rich). The bicycle continued to be more available to the masses for quite some time after.

The only twisting going on is in your comment, I'm sorry. :)

self-driving cars and lane splitting

Posted Feb 23, 2013 21:14 UTC (Sat) by khim (subscriber, #9252) [Link] (7 responses)

Regardless of exactly date very early prototypical automobiles were first made, it's utterly irrelevant to the question of the transport commonly available to people.

Ok. So now we redefine terms to mean "transport commonly available to people". Got that. No problem, you still lose: typical daily distance for novices in bicycle touring is 50km and unless you are professional (which should be excluded by your definition as irrelevant aberration) weekend tour distance is recommended to be 80-100km. Pretty close to this 72km stretch which was discussed above, isn't it?

The bicycle was in mass use *long* before cars were ever produced commercially (and then, in small numbers to begin with, available only to the very rich). The bicycle continued to be more available to the masses for quite some time after.

Depends in the country, really. China and Vietnam used bicycles extensively in XXI century (that means: they still use them today), but in US they were only popular as a tool for transportation for a couple of decades at the end of XIX century. 72km was, indeed, something you needed the whole day to travel back then (bikes from light allows and paved roads everywhere was not something which was common back then) and it's still not something people in poor countries do twice per day on bicycle.

Technology which makes bicycle capable to making at least couple of 72km trips per day don't come before car no matter how you twist it.

self-driving cars and lane splitting

Posted Feb 23, 2013 21:45 UTC (Sat) by paulj (subscriber, #341) [Link] (6 responses)

I'm not redefining anything. My original comment referred to "people", the original comment in the article referred to "our ancestors". To interpret the context as anything other than what applied to people living back then in a broad, plural sense is unreasonable, and only one someone trolling for an irrelevant argument might make. ;)

Yes, you're correct there were prototypical cars a long time ago, before even bicycles - slightly interesting, thanks. They were though very slow, impractical and very very *very* *few* in number - in no way available to people generally, such as "our ancestors".

Feel free to continue arguing that point though, I won't respond. ;)

As for distance, 72 km is well within the capabilities of many people, with a limited amount of training, from a few months to maximum a year (depending on pre-existing fitness). Humans actually have remarkable endurance. It's one of the few physiological characteristics we excel at in the animal world (least, amongst mammals), other than brain power. We used to survive by chasing, nominally faster, animals to exhaustion. Distances of 160 to 200 km are easily achievable within a day by average humans, on bicycles on paved roads, with just a year or so of adaption to the bicycle, as long as they eat underway. They won't do it at the speed of professional cyclists, but they can do it none the less.

As a data-point, my great-auntie used to cycle 40 km to find food during the dutch winter famine in WWII. On a bicycle with no tyres - which is far far more inefficient than with tyres. She had before then been no great cyclist (no more than normal dutch), never mind being a professional cyclist, and she would have quite under-nourished. Also, go to any road cycling club to find examples that will prove what I'm what telling.

Given your comment about paved roads, it seems you havn't actually read my comment. My speed and time estimates were for *unpaved* roads. With paved roads, sustainable speeds for average humans *increase* to 20 to 25 km/h and journey times for 72 km go down to under 3 hours.

That's definitely not all day. Indeed, it's short enough you could go early, get some business done, and make the return journey (though, probably better to stay the night).

A fit, but average, man on a bicycle can do distances in a day that would kill a horse, if you tried to make it do the same.

self-driving cars and lane splitting

Posted Feb 24, 2013 13:09 UTC (Sun) by khim (subscriber, #9252) [Link] (4 responses)

As a data-point, my great-auntie used to cycle 40 km to find food during the dutch winter famine in WWII.

Was it 40 km one way or was it 40 km + 40 km? Even if later (80km/day journey) it's still pretty close to one 72 km journey per day.

That's definitely not all day. Indeed, it's short enough you could go early, get some business done, and make the return journey (though, probably better to stay the night).

Please read again the words you wrote (I've highlighted them). If it's "better to stay the night" then it means 72 km journey took the "whole day". If you want talk in layman terms then "whole day journey" is something which can only be done once per day.

A fit, but average, man on a bicycle can do distances in a day that would kill a horse, if you tried to make it do the same.

Yup - but 72km is about what you can reach in a day. You can try to push for two such journeys in the same day if you really are desperate but for that you need paved roads (which come after car because they were mostly built for cars) and way-above-average-driver. Which returns us back to "72km journey takes the whole day".

self-driving cars and lane splitting

Posted Feb 25, 2013 6:32 UTC (Mon) by paulj (subscriber, #341) [Link] (3 responses)

The comment in the article, and my replies, are explicitly about the time taken to travel 72 km. Again, that's well within the capability of an average, mildly fit human, who otherwise doesn't cycle particularly far otherwise.

Such people probably wouldn't feel comfortable doing the round trip, and would likely be more comfortable staying the night. Someone with a bit more experience of cycling, but still otherwise ordinary and average, could complete the round-trip within a day. I imagine that, during the cycling boom in the late 19th and early 20th centuries, prior to the prevalence of the car, that that would not have been an insignificant proportion of the population.

The round-trip probably not being comfortable for those not adapted to cycling in *no way* changes the fact that the original comment in the article, that just the 72 km journey alone would take all day, is not quite right. Indeed, establishing exactly how far you can go in a day without a car wasn't even the main point of my comment. Rather it was meant to be a comment on the practicality and sustainability of the car as mass personal transport in a world where we increasingly live in more densely populated urban areas. Beyond some density of ownership and use, the car ends up being *slower* than many other forms of transport.

You've managed to derail that, already some-what off-topic, point into a fairly pointless nitpickery through shifting the goal-posts of context. Well done.

self-driving cars and lane splitting

Posted Mar 5, 2013 11:07 UTC (Tue) by emj (guest, #14307) [Link] (2 responses)

There might have been nit pickery, but it was an interesting discussion. In my view with paved roads biking 74km per day is easy, but 160km per day doesn't feel like something I would like to do more than once or twice a week. The problem with going longer distances in a straight line might be that the roads that are best for that kind of travel might be too car oriented.

self-driving cars and lane splitting

Posted Mar 5, 2013 22:19 UTC (Tue) by dlang (guest, #313) [Link] (1 responses)

paved roads are primarily a car legacy, paved roads are worse for livestock than unpaved roads as well as being very expensive to build and maintain.

Yes the Romans built paved roads, but they built them to march their infantry along, not to help commerce.

self-driving cars and lane splitting

Posted Mar 6, 2013 7:36 UTC (Wed) by paulj (subscriber, #341) [Link]

Modern paved roads were lobbied for and started to be built for cyclists, before the motor-car existed (other than some very rare, prototypical mechanical carriages): http://www.roadswerenotbuiltforcars.com/

Energy efficiency

Posted Feb 27, 2013 23:58 UTC (Wed) by kleptog (subscriber, #1183) [Link]

I have to agree with you that human can travel amazing differences on a bike. Even living in Holland you can surprise people by say you can travel 50km in two hours. I've also done cycling trips and figured out that you can average 10km/h over a whole day including all breaks for eating without even really trying.

But one thing I found interesting was that if you look up energy efficiency of travel for animals (in Joules per km) human are fairly average and eagles do the best. If you you include humans on a bike then we handily beat every other animal. Bikes are a very efficient form of transport.

The only reason modern bikes took so long to invent is that the materials for needed for the gears and chains (also needed for cars) just didn't exist.