Aerospace & Engineering

Learning About Rocket Science, One Small Step at a Time

Kumar Venkatramani · September 2026

On September 28, SpaceX's Starship reached orbit for the first time, on its fourteenth test flight. The headlines were about the milestone. The question my inner child asked was much simpler: how does a rocket actually steer in space?

It's a deceptively simple question. In a car, you turn the steering wheel, a linkage angles the front wheels, and as the engine drives the car forward, the grip of the tires on the road carries it around the turn, in the direction you chose with the steering wheel.

A rocket has no wheels and no pavement. So what exactly does it steer with? Since it flies through the air, maybe the way an airplane steers holds a clue.

In an airplane, the pilot moves the control wheel or stick, which tilts hinged panels on the wings and tail. The air rushing over those panels rolls the plane into a bank, and the lift from the tilted wings pulls it around the turn.

But a rocket doesn't have wings either. Some rockets do carry fins, but fins need air, and a rocket spends much of its climb where the air is too thin to help, and all of its time in space with no air at all. Yet it has to bend its path into just the right arc to reach orbit, and turn itself around in the vacuum of space. Chasing that question sent me down a rabbit hole.

An outboard motor for the sky

It turns out the answer is elegant. Think of a small boat with an outboard motor and no rudder. To turn the boat, you swing the whole motor. Its push, which drives the water out behind it, now pushes the back of the boat sideways, and the boat turns. A rocket works on the same principle. Its engine pushes hot gas downward with enormous force, and that pushes the rocket up. Tilt the engine slightly, and its push swings the rocket around to a new direction, as Figure 1 shows.

Most large rockets don't have a single engine, though. Starship's first stage, called the Super Heavy booster, has 33 engines arranged in three rings (Figure 2). The outer twenty are fixed in place. The inner thirteen sit on a pivot mount called a gimbal, the same idea as the gimbal that keeps a phone camera steady while you walk. Tilt an engine a few degrees and its push no longer passes straight through the rocket's center of mass, so the rocket rotates toward the direction you want (the same effect as in Figure 1). Engineers call this thrust vector control, which simply means controlling the direction of the thrust.

What about the upper stage, and other ways to steer?

The upper stage (also called Starship, or "the ship") has six Raptor engines: three tuned for use in the atmosphere, which gimbal, and three with large nozzles for the vacuum of space, which are fixed.

Technically, engines aren't the only steering tools. Coming home through the atmosphere, the booster also steers with large lattice-shaped grid fins near its top. In space, with its main engines off, the ship turns using small gas thrusters. But to a first order, pivoting the engines is how a rocket steers.

The next question is: who pivots the engines, and how? Starship flies without a crew so far, so the answer is the flight control software. But the engines are heavy, and each one may be producing more than two hundred tons of thrust. Swinging one while it fires, precisely, within fractions of a second, again and again, while the whole vehicle shakes its way through the atmosphere, takes an enormous amount of force. Where does that force come from?

Borrowing from airplanes

For most of the history of spaceflight, the answer has been hydraulics. As an electronics engineer, hydraulics sits outside my own training; it's home ground for mechanical engineers. So for those like me, here is a simple explainer.

On a small plane, when the pilot turns the control wheel, steel cables and pulleys physically move the panels on the wings and tail. On a large jet, the forces get too big for that. At 500 miles an hour, the air pushing against those panels is far more than a pilot's arms could comfortably overcome for hours on end. So on an airliner, the pilot's input doesn't move the panel directly. It opens a valve, and fluid under pressure does the grunt work.

The principle is one most of us met in high school physics as Pascal's principle. Picture two syringes, one small and one large, joined by a tube and filled with fluid. Press the small plunger gently, and the large one pushes out with far more force, because the pressure is the same everywhere in the fluid but acts on a larger area. A small push, multiplied. Airliners run their hydraulic fluid at around 3,000 pounds per square inch, and use it to move the control surfaces, extend the flaps, raise the landing gear and apply the brakes. Most large airliners carry three or more independent hydraulic systems, so that if one fails, the others keep the airplane flying.

Rocket engineers borrowed the same idea to swing their engines. The massive F-1 engines on the Saturn V rocket that carried the Apollo missions to the Moon were steered hydraulically, and so were the Space Shuttle's main engines, powered by dedicated units that existed only to run the hydraulic pumps. For about seventy years, that is simply how rockets were built.

Hydraulics comes with baggage, though. You need the fluid, the tanks to hold it, the pumps to pressurize it, the lines to carry it and the pistons at the end. The pump works to keep the whole system pressurized whether anything is moving or not. And wherever there is fluid under pressure, sooner or later there's a leak. Anyone who has owned an older car has seen this: a small reddish puddle of power-steering fluid, or a soft brake pedal from leaking brake fluid, since both systems run on hydraulic pressure. On an airplane, leaks are why there are three systems. On a rocket full of fuel and liquid oxygen, a leak is a fire risk.

Starship goes electric

Now picture all of that plumbing, thirteen times over, on the booster. SpaceX decided to get rid of it. Cars had already shown the way: most new cars, gasoline or electric, have replaced the hydraulic power-steering pump with an electric motor that helps only when you turn the wheel.

On Starship's booster, there are no hydraulic pumps, no hydraulic fluid and no high-pressure lines. Each steering engine is swung by an electric motor turning a precision screw, much like the screw jack you'd use to raise a car while changing a tire, where turning the handle slowly lifts the whole car. Here the motor does the turning, fast and with great precision, and the screw pushes a two-hundred-ton-thrust engine exactly where the flight computer wants it. The power comes from lithium-ion batteries housed in the long fairings running down the side of the booster. SpaceX says the new system "has fewer potential points of failure and is significantly more energy efficient than traditional hydraulic systems." It first flew on Starship's second flight, in November 2023.

To give credit where credit is due, SpaceX wasn't the first to do this. Europe's Vega rocket has steered its upper stages with electric motors and lithium-ion batteries since its first flight in 2012. Rocket Lab's Electron goes further still, using battery-powered electric motors to pump propellant into its engines, a job that in most large liquid-fuel rockets is done by turbines driven by burning some of that propellant. What SpaceX has done is carry electric steering to a scale nobody had attempted.

The same shift in cars, airplanes and rockets

Machine	Hydraulic (or mechanical) way	Electric way
Car steering	Engine-driven pump, fluid, hoses	Electric power steering motor
Airliner	Air bled off the engines, larger hydraulic systems	Boeing 787: electric brakes, electric compressors for cabin air, smaller hydraulics
Fighter jet	Central hydraulic system to flight controls	F-35: self-contained electro-hydrostatic actuators (a motor and a small sealed hydraulic unit at each control surface)
Rocket steering	Hydraulic actuators on each engine	Vega (2012), Starship booster (2023): electric actuators
Rocket propellant pumps	Turbopumps driven by burning propellant	Rocket Lab Rutherford (2017): battery-powered electric pumps

The F-35 row matters. It's a hybrid that keeps a little hydraulic fluid inside each actuator. The transition doesn't have to be all or nothing.

Why now, and what it buys

I think the answer to "why now" is that the ingredients finally got good enough. For decades, hydraulics won because motors and batteries were simply too heavy for the force they delivered. Then three things improved dramatically, driven largely by consumer electronics and the electric vehicle industry: lithium-ion batteries that store far more energy per kilogram, compact motors built around powerful rare-earth magnets, and power electronics that switch large currents efficiently.

What they buy is more than weight savings. SpaceX's own claim is that the electric system is significantly more energy efficient, in part because there is no pump working to keep a whole fluid system pressurized. There is no fluid to leak and nothing to top up between flights, which counts for a lot on a vehicle meant to fly again and again. And an electric actuator (the part that does the pushing) knows exactly where it is and how hard it is pushing at every instant, and reports that straight to the flight computer. Steering stops being a plumbing problem and becomes a control problem, which is exactly what computers are good at.

How much weight? I've seen figures of roughly a ton per engine, or 13 to 15 tons across the first-stage booster. I couldn't find SpaceX confirming any number, so I treat that as an estimate. I'm confident about the direction; I'm much less sure about the size. And a ton saved on the booster doesn't translate into a ton of extra cargo, for reasons in the box below.

What the weight savings are really worth

A ton saved doesn't mean a ton more cargo. A common rule of thumb: a kilogram removed from the upper stage, which rides all the way to orbit, adds roughly a kilogram of payload. A kilogram removed from the first stage, which drops away a few minutes in, adds only about a tenth of a kilogram, sometimes less. So even 15 tons off the booster is closer to one or two tons of extra payload. Savings on the ship count more, ton for ton.

One caution, because it would be easy to overstate this. What's being replaced is hydraulics, fluid under pressure used to move things. Fluid dynamics isn't going anywhere. The Raptor engine itself is a masterpiece of it: how the propellants flow, how they burn, how the nozzle is shaped. The fire is still very much a fluid problem. It's only the steering that has gone electric.

Four takeaways from Flight 14

Flight 14 gave the electric steering a demanding day.

First, engines failed on both stages. The booster lost one engine on the way up, and the upper stage lost one of its six. When an engine quits, its push vanishes from one side, and the remaining engines have to tilt to rebalance the vehicle. The steering did its job, and the upper stage still reached orbit. I wouldn't claim electric steering is why it survived; having many engines and good flight software does most of that work, and hydraulics could have done the tilting too.

Second, the payload was delivered. Despite the engine loss, the ship released 26 of SpaceX's new, larger Starlink V3 internet satellites into orbit, three of them carrying cameras to photograph the ship's heat shield.

Third, the booster came back. It flipped around and relit its engines to head back toward Texas, lost a second engine during that burn, and still made a controlled splashdown in the Gulf of Mexico.

Finally, not everything went to plan. The flight was meant to last about ten hours and six orbits. After the engine loss, SpaceX brought the ship down early, and it splashed down north of Hawaii about three hours after launch. SpaceX's approach of testing in public, fixing and flying again has its critics, and a mission cut short by two-thirds gives them something to point to. Reaching orbit with a real payload is still a genuine milestone.

Where a simple question leads

Engineering is always a problem of trade-offs. SpaceX gave up a technology with seventy years of flight heritage and took on motors, batteries and electronics that have to survive brutal heat and vibration. In return, it got a lighter, simpler vehicle with fewer things that can leak.

My inner child's question was how a rocket steers. The answer turned out to be a window into something much larger: fluid power, one of the great workhorses of mechanical engineering, steadily giving way to electrification, first in cars and airplanes and now in the largest rocket ever flown. It really is rocket science. It's just increasingly electrical.