While the colonization of the Solar system is underway, a new sport emerges: Spatial Rowing.

The rules are the following:

  • Each spacecraft has one driver (the contestant)

  • The loaded mass of each spacecraft (defined as rest mass including the driver, its spacesuit, and the propellant) may not exceed two metric tons.

  • The linear size of the spacecraft may not exceed 4 meters in any direction.

  • Any form of stored energy is forbidden on the spacecraft, except energy stored inside the driver (in its muscles, fat...) Of course, this is impossible to literally comply, so the following subrules exist:

  • No antimatter, fission and fusion fuel.

  • No gas under pressure exceeding 1000 Pa (at launch), except the breathing mix of the driver.

  • The breathing mix cannot be used as cold gas thruster, except if it is directly blown out by the driver.

  • Temperature difference exceeding 20 K between two parts of the craft (at launch) is forbidden.

  • No solar panels or beamed power receivers.

  • At launch on the craft may not have any two chemical substances which produce exothermic chemical reaction with each other with ignition temperatures between 0-600 K, except if both are inside the living, organic tissue of the driver, or one is inside the driver, and the other is oxygen (for life support) or nutrients for the driver.

  • No temperatures (during the whole race) exceeding 600 K are allowed.

  • Burning life support oxygen and nutrients externally to the driver is forbidden during the race.

  • No spinning flywheels and parts under tension or torsion exceeding 10 J are allowed(at launch).

  • The spacecraft have to have a small accumulator, for providing emergency transmission power and life support power, if something should happen to the driver, but using it means giving up the race.

So in one word, the craft must be entirely man-powered, including life support and propulsion. (Of course, there can be accumulators on board, but charging them is to be done by the driver.)

The objective of the race is to go through a racetrack, situated in the space, which is some kilometers long, and contains obstacles and momentum exchange wheels. A pre-set trajectory should be followed, and the obstacles not touched. A typical racetrack takes about one hour and 50 m/s delta V (minimally)

The aim is to do it with the minimum amount of faults, and in minimal time.

What kind of propulsion system should such a racing craft have?

  • 1
    $\begingroup$ oars would be an option $\endgroup$
    – Kilisi
    Jun 4, 2017 at 7:10
  • $\begingroup$ I suspect the best option is to have some form of turbine powered by a pedal and gear system to generate electricity. $\endgroup$ Jun 4, 2017 at 8:07
  • $\begingroup$ Does the course run mostly near planets? $\endgroup$ Jun 4, 2017 at 8:15
  • $\begingroup$ @Bellerophon The race takes place on high orbit around moons, but since the whole course is only some kilometers long, orbital mechanics don't come into play, and the race is never organized in the presence of strong magnetic fields. $\endgroup$
    – b.Lorenz
    Jun 4, 2017 at 10:59
  • 1
    $\begingroup$ Ok, here are some suggestions: 1) make it "downhill" towards some planet/moon so gravitation kicks in. 2) allow for a starting momentum, e.g., let your athlets push themselves from the starting block for example. 3) allow for maneuvering thrusters that do not give you forward momentum but allow you to steer ( be careful, exploitable) 4) be ok with your race taking a couple of months $\endgroup$
    – Raditz_35
    Jun 4, 2017 at 11:29

6 Answers 6


I suggest a mass driver, using counter-rotating centrifugal slug throwers. The rotating components are mechanically geared, essentially using a bicycle crank approach.

The counter-rotating part, obviously, is necessary to maintain some sort of stability during the pedaling portion of the exercise. As a bonus, differential braking of the mechanism will, in moderation, allow using it as a momentum wheel and controlling attitude, at least in the plane of rotation.

Now, for numbers. "some kilometers" is not what one calls a useful number, since it can refer to pretty much any distance from 1 to 100 kilometers (or more).

Assuming a "racetrack" involves an outgoing velocity of x m/sec, with the same return velocity and little lateral maneuvering, a delta-v of 50 m/sec suggests a velocity on the order of 15 m/sec, or (roughly) 1 km/minute, for a nominal race time on the order of 2-3 minutes per kilometer of racetrack overall length. Assuming this is intended, a race time of 1 hour implies that "some kilometers" is on the order of 20 km. Note that this implies relatively high accelerations during such phases, in order to keep the acceleration times as short as possible.

It is also obvious (or it should be) that placing an upper limit on vehicle mass is pretty silly, since nobody in their right mind will carry more mass than necessary. So let's work with a total mass of 1000 kg. We'll also assume an essentially massless centrifugal apparatus, in order to limit wasted energy in unused rotational energy. Furthermore, just as a starting point, figure on 300 kg for the total slug masses in each phase, for a total of 600 kg and 400 kg for pilot, life support, frame and mechanism. Finally, let's set the radius of the slug mechanism at 1 meter.

At the starting gun, the pilot pedals for an effective power input of 300 watts for 20 minutes and ejects the first mass. He then coasts for another 20 or so minutes (pedaling furiously) and then ejects the other, reversing course. Does this work?

20 minutes is 1200 seconds, for an energy of 360,000 J. Rotational velocity is about 25 m/sec. So the ejection momentum is about 7500 kg-m/sec, and the delta v of the ship is about 10.7 m/sec.

Coasting for 20 km will take about 1870 sec, or 31 minutes, which is borderline encouraging.

The energy of the remaining slug will be 180,000 J, and in 1870 seconds an additional energy of 560,000 J will be added, for a total of 740,000 J. Ejection velocity of the second mass will be about 50 m/sec for a momentum of 15000 kg-m/sec. Since the release of the second mass represents a larger mass fraction, delta-v will be about 37.3 m/sec. This results in a reverse velocity of 26.5 m/sec, and the return leg will take about 750 seconds, or about 12 minutes.

So, ignoring other maneuvering, total race time is about 63 minutes, which seems a pretty close fit to the stated requirements. Total delta-v is 49 m/sec, much to my surprise.

Of course, this requires the pilot to produce 300 watts for 50 minutes, but this is well within published norms (400 watts for 1 hour). It also implies a perfectly efficient mechanism, but let's figure this efficiency loss is within the output capability. It also implies a fairly high release recoil. This is to some degree mitigated by the fact that slug release will need to be in two separate stages on each wheel, in order to avoid an out-of-balance condition for any length of time, so the return-phase recoil will be about 20 m/sec in two stages, closely spaced. This will clearly require pretty good shock absorbers, but this is left as an exercise to the reader.

ETA - A little thought shows that the shock absorber issue is no big deal. Rather than releasing the 300 kg in two (balanced) masses of 150 kg, it's obvious that a much larger number of smaller masses can be used. For instance, if 300, 1 kg masses are employed, the delta-v per release will be on the order of .16 m/sec. This can obviously be spread out over several revolutions, providing a much gentler thrust event.

  • 1
    $\begingroup$ The larger number of smaller masses doesn't just provide a gentler thrust event, it also gives more opportunities for the pilot to correct his course as he goes. In addition, it would allow for more things to happen during the course of a race for the spectators, which is a consideration if such a sport were to ever catch on (I still think it would be pretty boring to watch though: it would make sailing look like NASCAR). $\endgroup$
    – JBiggs
    Jun 5, 2017 at 15:53
  • $\begingroup$ Larger number of smaller masses will also allow to accelerate earlier. For a big chunk you need to wait till the power is accumulated. For smaller chunks your can push the first one almost immediately. $\endgroup$
    – keiv.fly
    Jun 8, 2017 at 0:32
  • $\begingroup$ I would have a big tank of water and use pedal power to drive a pump rather than throwing solid slugs. I think this would allow the 'driver' a fair bit of control over the mass ejection direction (move the nozzle), though would likely have more loss than a pure mass driver. $\endgroup$
    – sdfgeoff
    Sep 16, 2017 at 7:10

There will be a bicycle in a spacecraft. You push the pedals, generate electricity and use it for the engines (for example, ion engines).

Also pedals connected to a flywheel or just turning a flywheel will directly change the direction of the spacecraft. It will rotate in the opposite direction where the flywheel is rotated.

People can generate around 1000 Watts (surpassing the microwave oven) for 30 seconds. See link here: https://bicycles.stackexchange.com/q/21294

  • $\begingroup$ 1000 Watts and all is nice and I think this idea would work, I advice the author: If you choose this option, you should do a bit of the math behind it (I'm not suggesting I can). One would need to power several devices - you need a magnetic and electric field, some valves to open when needed and much more - and see how much thrust remains at the end from the emitted particles. I'm sorry if I understood the answer incorrectly however $\endgroup$
    – Raditz_35
    Jun 4, 2017 at 8:46
  • $\begingroup$ Oh PS: Maybe difficult to implement if you want less than 600K at all times $\endgroup$
    – Raditz_35
    Jun 4, 2017 at 9:00
  • 1
    $\begingroup$ I suggest looking at e.g. space.stackexchange.com/q/8599/415 for some real-world data on ion engines. If I'm reading space.stackexchange.com/a/8601/415 correctly (I'm too lazy to do a lot of research), NSTAR needs 2.3 kW for 30 mN of thrust. $\endgroup$
    – user
    Jun 4, 2017 at 13:19

After the claims of Raditz_35, that it is not possible, I have actually looked into the possibilities of man-powered spacecraft. My solution is a crossbow-bike:

The driver either sits in a thin, inflated bubble or wears skintight spacesuit, and pushes the pedals (linear if possible, to avoid spinning up the craft.) His/her work is used to bend a bow or expand a rubber, which is then used to launch a small dart of solid fuel.

A cyclist is said to be able to provide 400W for more than a hour, but since it wears a (super-duper-sci-fi) skinthight spacesuit, and there are losses through the mechanichsms too, let's count with 300W.

The propellant has kinetic energy of m*v** 2/2, and momentum of m*v, so the thrust is 300*2/v. Even a very bad crossbow can fire at 50m/s, so the thrust is 12N. If higher acceleration is needed, they can gear down, and, while loosing Isp, have thrust as high as 100N. Assuming that the craft and the driver are 250kg (human, bubble/suit, oxygen, drive mechanism), an 1000 kg spacecraft with mass ratio of 3 has about 54m/s of delatV. It's acceleration is a pathetic 0.012 m/s** 2 at launch, but it gets a little better as the propellant runs out. The attitude control is achived by shifting weight, and thus making the thrust off-axis. (great skill needed)

  • $\begingroup$ Efficiency is a huge concern here. A crossbow's arms move as it shoots. Those arms require kinetic energy to move, and that energy dissipates and does not turn into thrust. Whatever your firing mechanism is, the mass and speed of the projectile should be as large as possible compared to the mass and speed of the other parts. $\endgroup$ Jun 5, 2017 at 4:02
  • $\begingroup$ That sounds super dangerous in a space race. I would still go with my new air-pump solution, but I do agree that this would work $\endgroup$
    – Raditz_35
    Jun 5, 2017 at 5:58
  • $\begingroup$ @BobTheAverage Quite right, there is a great exercise for the engineers of the teams. But probably they can reduce the moving mass quite much with modern elastic materials. $\endgroup$
    – b.Lorenz
    Jun 5, 2017 at 6:09
  • $\begingroup$ @Raditz_35 If the contestants are on the racetrack simultaneusly, there is geat danger, that ones solid propellant raptures the others bubble. But if they make their try after each other, there is no danger. Still, the pump solution is viable too. The theoretical maximum Isp*thrust of both systems is constrainted by the max power output of a human, but the liguid/gas pump might has better efficiency. $\endgroup$
    – b.Lorenz
    Jun 5, 2017 at 6:15

After some iterating, I decided to feature my comments in an answer because I feel like they are legitimate enough.

If you do not know this, google or youtube conservation of momentum. The thing is: If you want to move forward, you need something to leave your spacecraft which travels the opposite way. So here are ways the human body can do this:

1) Farting. Yes, very funny - imagine if one dared to write make-believe which wasn't dead serious. Some people practice this - check out flatulists. There exist variations on this for women. I'm serious, culturally you might think it's disgusting, but it is an artform to some and historically, people weren't as prude. The good thing is that the driver can face forward.

2) Sneezing. The fastest way of exhaling. Who knows what can be done with some practice. Of course this is very bad for your health, but oh well, that's how sport works.

3) Just blowing out air. We do it constantly. It would be enough - eventually.

In order to implement this, you of course need a valve of some sort. Maybe build some (relatively low) pressure first or let it directly slip into the universe.

4) Oh, yes, I first thought of farting before that: Just pump out air via a manual pump. This is easy enough.

If there were things along the racetrack for the athletes to push themselves forward with, things become a lot less disgusting - but you would have to allow some sort of stationary objects along the way to push forward from.

If you would do this on a moon with an atmosphere, things would be A LOT easier. It might even be a better story because moons are awesome.

  • $\begingroup$ The race takes place in Space. There is vacuum, you can not expose the driver to it (to make him sneeze or fart). It is forbidden to touch the obstacles (altough there are objects on the racetrack.) You have to come up with a propulsion system powered by humans. $\endgroup$
    – b.Lorenz
    Jun 4, 2017 at 11:07
  • $\begingroup$ @b.Lorenz You do not have to expose your driver to vacuum to implement this. I thought it was so straight forward that I chose not to comment on this, but I added an explanation $\endgroup$
    – Raditz_35
    Jun 4, 2017 at 11:10
  • 2
    $\begingroup$ I would like to add forceful urination to this list. Loss of breathing gas is prohibited (so sneezing out but farting still ok). There is no cap on loss of liquid. After me and my suit that is space for a lot of beer, which I think is called "propellant" in bullet point 2. The beer companies can sponsor. $\endgroup$
    – Willk
    Jun 4, 2017 at 11:29
  • $\begingroup$ @Will The gas rule is newer than my post. But nobody says you have to breathe in and sneeze out breathing gas. Could be plain nitrogen and you breathe inbetween. But you are right, urination would build quite a thrust - but unlike my 3 options, you are pretty limited with how much you can urinate. Would be fun though to do a piss-dog-race in space for example $\endgroup$
    – Raditz_35
    Jun 4, 2017 at 11:32
  • $\begingroup$ I have prohibited loss of breathing air, since then you could pack gas saying that it is breathing mix, and since it has to be at high pressure, use it as cold gas thruster. But loss of liquid, as long as it is propelled by the driver's muscles (like pee), is definitely ok. $\endgroup$
    – b.Lorenz
    Jun 4, 2017 at 11:34

Considering the law of conservation of momentum, barring any external sources of power (e.g. solar/magnetic sails, lasers, etc), the only way to push something forward is to expel something backward. Combining with @Bellerophon's comment, that the best possible way to be for the human to work on producing electricity, which would then be used for an electric propulsion system (https://en.wikipedia.org/wiki/Electrically_powered_spacecraft_propulsion).
The constestants would have an exactly specified amount of propellant available. All other life support systems would be on separate power and batteries, their weight fixed and properly accounted for. The goal would be to get to the end space coordinates as fast as possible, Various obstacles vould be interspersed between start and finish, like asteroids, magnetic irregularities, etc.
No matter the details, in the end it will be a contest of how much electric power can the human provide, in addition to maneuvring dexterity and piloting skills.
(Kudos to @Raditz_35 for the ideas of farting and sneezing and blowing air, they might actually work!)


Solar Sails.

Without any sort of compressed gas, chemical reactions, or flywheels, the only way you're going to get a spacecraft to move is via something outside. A human simply cannot provide any means of propulsion - Or, for that matter, attitude control - for a vehicle in space. All earth-based vehicles rely on the human pushing against the ground, water, or air - Of which there is none.

A solar sail powered ship would rely entirely on how the sail is set by the pilot. A handful of different sails would allow for maneuvering (Terribly and ineffectively, but possible), and the propulsion is not included in the ship.

They would be incredibly slow and incredibly ungainly. Any sort of docking or undocking would have to be done via some form of small tug/tender. They're not going to be precision craft - Any sort of obstacles would have to be measured in thousands of kilometers at the very least, and the timeframe would likely be weeks or months. Maybe even years. Solar sails are slow.

  • $\begingroup$ This is excluded by the OP. But, I have read fiction where manual craft were sailed for sport. $\endgroup$
    – JDługosz
    Jun 4, 2017 at 9:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .