In May 2013, things were not looking good for the Juno mission, at least to the untrained eye. Only halfway to Jupiter, the Sun was already starting to pull it back towards the center of the Solar system.
With a low-powered Atlas V rocket which had just 1/10th the carrying capacity of the Saturn V used to reach the Moon, you would think that the mission planners had made a huge mistake.
Over the next year, anyone watching would have seen the doomed spacecraft coming extremely close to Earth, as if it was on course to becoming another piece of flaming debris falling through the atmosphere. Not what anyone would want from a 1.1 billion dollar project.
What happened next would be called a miracle if not for the fact that it had been planned that way from the start. The spacecraft missed crashing into South Africa by less than 100 kilometers, nearly doubled its speed, and then spent two years hibernating before waking up three weeks ago to prepare to enter Jupiter's orbit.
This is the secret to why modern missions with robots are possible. Such missions have budgets only a fraction of that used to send humans on the Moon.
The secret is to do multiple passes close to the planets in the solar system to pick up speed for the next part of the journey.
The technical term for these maneuvers is "gravity assists." The trade off is time for fuel. By meandering around the solar system, picking up speed wherever the planets align to allow it, missions can cut their fuel needs by a factor of 10 or more. In fact, they can cut fuel needs by a lot more in some cases.
Juno only used one gravity assist, as it passed close by the Earth. Without this, the spacecraft would have had enough momentum only to reach the asteroid belt.Many contemporary missions use this sort of strategy, and in fact, the trajectory of Juno was comparatively simple. Juno had a much simpler trajectory than Cassini, which was meant to go to Saturn. Cassini took a total of four detours: two detours through Venus, one through Earth and one through Jupiter. Galileo passed by Venus once and Earth twice before reaching Jupiter. Both then did an even more intricate dance around the moon systems of the giant planets.
The field of orbital dynamics has been developing in leaps and bounds and has become increasingly "dynamic" since the 1980s.
Before this, the maneuver used to get from one orbit to another was called a Hohmann transfer. Unlike the more exotic maneuvers, the plans for a Hohmann transfer can be worked out using pen and paper. They are predictable, repeatable and simple to plan for. The downside is that they require huge quantities of fuel. This leads to missions having to reduce the number of scientific instruments they carry.
Modern maneuver planning, by comparison, combines machine learning and chaotic systems. Unfortunately, such problems are also some of the hardest to solve. The usual method of machine learning, called gradient descent, fails miserably. The "solution space" in which trajectories live is incredibly sparse and complex. Changing the height of an orbit which is 940 million kilometers long by as little as 10 kilometers will mean the difference between crashing into the Sun or escaping the solar system entirely. This comes from the fact that orbits are inherently chaotic. The three body problem -- or Sun, Earth, Moon system -- was the very first example of Chaos in classical mechanics before the word was even coined.
Most missions today analyze as many "slingshots" as possible but each takes time. A lot depends on what you optimize for, and there is never a guarantee that you will find a promising minimum, even a local one. Conflicting mission goals also make planning very difficult.
Newer methods don't just use stable parts of orbits around planets to plan the slingshot, but also unstable and semi-stable points of the system, the Lagrange points. There are even more exotic methods for very long term missions called low-energy transfers, where energy savings are possible even when there are no bodies between the launch point and destination point (an example would be the Earth Moon system.)
The field is new and heuristics-based, and will only grow in importance. Unfortunately, we have no idea just how much more we can shave off of the fuel requirements for orbits. The 2-solution spaces can't be solved analytically, and numerical methods choke on how difficult the spaces are.
Juno settled into orbit around Jupiter this week, according to the Guardian. The Verge notes that NASA named Juno after Jupiter's mythological wife. The gag is that Juno has now been sent to "check up" on Jupiter.
NASA-critics took up the gag and chided NASA for their cheekiness. The gravitational pull of the mighty Jupiter sucks in comets and asteroids and thus protects earth from debris that "could really rock us" if it managed to hit planet Earth.
[Jupiter] is known as Earth's "comet vacuum cleaner" or "comet liver," because its enormous gravitational pull sucks in comets and asteroids that could really rock us if they made it here. One Reddit commenter points out that NASA's scheming to get Jupiter busted seems sort of rude. And it's true — it's not a very kind way to repay all of Jupiter's years of loyal service protecting us from possible extinction events.
[Photo by David McNew/Getty Images]