Wikipedia:Reference desk/Archives/Science/2016 August 22

= August 22 =

probabilistic sampling path planners
Besides Probabilistic road map and rapidly growing tree (RRT), what other algorithms fall under the probabilistic sampling category of path planners in robotics?117.206.48.169 (talk) 14:29, 22 August 2016 (UTC)
 * I would put any statistical estimator, like a Bayesian state estimator or a Kalman filter in that category.
 * There isn't a "correct" taxonomy of algorithm styles; different authors and researchers have different insights into the problem-space, so they may place various methods in the same category for a variety of reasons.
 * So - why do you need to categorize these algorithms in the first place? Are you looking for comparisons and contrasts?  Reference implementations?  Practical examples or case-studies?
 * Nimur (talk) 16:28, 22 August 2016 (UTC)

Extreme track and field
How fast could a human run 100 meters with a 30 mph tailwind and the fastest reasonable track profile? Let's say that slope cannot exceed 45° anywhere and the human must use a running motion the whole way and stop without assistance or falling down. A slope change to flat or uphill to aid stopping is allowed. Distance is measured by how far the center of mass goes cause 1 meter of steep slope is only ~0.7 meters horizontally. Sagittarian Milky Way (talk) 23:14, 22 August 2016 (UTC)


 * Given the effect of a steep slope (and of a strong wind) on a human's balance, my suspicion is that it would probably slightly slower than Usain Bolt manages on the flat. Wymspen (talk) 09:10, 23 August 2016 (UTC)
 * Wouldn't that mean a moderate or slight slope is fastest? 30 mph wind isn't that balance disrupting. I did this in Hurricane Sandy (downhill) and literally could not slow down until I discarded the umbrella built for two. I stopped from being dragged into traffic without dropping the other (smaller) umbrella. The wind was probably at least 40. Sagittarian Milky Way (talk) 09:44, 23 August 2016 (UTC)
 * This article may help you with your research. -- Jayron 32 11:56, 23 August 2016 (UTC)
 * Double the speed means quadruple drag which requires eight times the power to overcome. (I may be off slightly) Drag equation I wonder if they took that into account.--Savonneux (talk) 13:08, 23 August 2016 (UTC)


 * There is a lot of research in the field of running and human mechanics. What you appear to want is the maximum speed at which a human can move his or her legs fast enough to keep up while applying enough downward force to remain in an upward position. If the legs do not move fast enough, the person will trip. If there is not enough downward force, the person's torso will get closer to the ground, causing a foot to drag, causing the person to trip. I am partial to Dr. Peter Weyand's work. Depending on conditions, he has placed the top running speed of a human between 40 and 50mph. The 40mph is technically possible for a very tall and genetically perfect running human (in other words, nearly all legs). The 50mph limit is based on the maximum speed a human can quickly move his or her legs back and forth as well as the time required to transfer a downward force from the leg to the ground. You can see a summary of biological running limitations here, but it isn't one of his better publications. 209.149.113.4 (talk) 18:35, 23 August 2016 (UTC)


 * A tailwind could certainly help, provided the runner could adjust to it. But I would think any significant downward slope would be counterproductive, as it would mess with the runner's balance. As we saw in some of the recent Olympic footraces, some of them had trouble staying balanced even on a flat surface. ←Baseball Bugs What's up, Doc? carrots→ 18:42, 23 August 2016 (UTC)
 * I think with practice, the downward slope would NOT be a hindrance. It would be a hindered to someone trying to run UPRIGHT, but the "trick" would be to throw yourself down the slope and "balance" to the ground you are running on, not to the horizontal, so that gravity effectively starts "adding" to your forward vector. The extreme example is how a skate boarder can thrown them self down a half pipe and essentially "balance" on a vertical wall on the way down. Vespine (talk) 22:27, 23 August 2016 (UTC)
 * This is definitely something that sources are needed for; on the other hand, some direct experience would be invaluable also. I can look through stuff like  but, well, it's like explaining sight to a blind man.  They suggest that runners putting in "even effort" will go faster downhill.  But definitely their bodies have to deal with more impact, because the gravitational potential energy is being dissipated in their tendons and muscles, essentially. Wnt (talk) 01:53, 24 August 2016 (UTC)
 * Pure anecdote, but in my experience of having a steep hill between my college and the train home, it's much easier to keep your balance by skipping than by running. Unfortunately, it's very difficult to slow down (or, in fact, to not accelerate) until you hit the flat, since the skipping motion doesn't permit any balanced deceleration, and transitioning to a run just makes you fall over.  MChesterMC (talk) 08:07, 24 August 2016 (UTC)


 * Skipping should slow you down. The first contact, per foot, pushes forward to create the skip. The second contact pushes back to accelerate. It is likely that you can achieve the same speed by just running. A benefit to skipping is the increase of vertical force. The first contact acts as a sort of pole vault, giving the torso upward velocity. As I mentioned above, from many decades of compiled running research, the two issues with high speed running are moving the legs back and forth (there is a maximum limit to how fast a human is capable of contracting and relaxing leg muscles) and maintaining enough distance between the torso and the ground (if 100% of energy was used for horizontal momentum, vertical momentum would be lacking and the person't torso would get lower and lower to the ground until running was not possible). 209.149.113.4 (talk) 19:01, 24 August 2016 (UTC)