Flying Machines

On 21 November 1783, two Frenchmen rose over Paris standing on the gallery of a cloth-and-paper balloon, feeding a fire to keep it up. Twenty-five minutes and about nine kilometres later they came down safely: the first people known to have flown freely through the air and lived to describe it. A hundred and twenty years after that, two bicycle makers from Ohio flew a powered aeroplane for twelve seconds. In 2025, airlines carried just under five billion passengers on more than 38 million flights.

This piece follows that arc, from people who jumped off towers to what the Wright brothers actually invented, how a wing really holds a 79-tonne airliner up, and the questions aviation faces now.

Part 1 — The dream before the machine: balloons, Cayley and Lilienthal

The oldest flying machines that really worked were toys. Kites may have been invented in China as far back as the 5th century BC; tradition credits the philosophers Mozi and Lu Ban. Chinese children were also playing with the "bamboo-copter", a stick with a little spinning rotor that climbs when you rub it between your palms and let go, from around 400 BC. Both contain a real principle of flight: air pushed one way pushes back.

People were less lucky than toys. The best-known early attempt is that of Abbas ibn Firnas, a scholar in Córdoba in the 9th century, who is said to have covered himself in feathers, strapped on wings, jumped from a height and glided some distance before landing so hard that he injured his back. He blamed himself for forgetting a tail. It is a vivid story with a problem: it comes from a single source, the historian al-Maqqari, who wrote it down around seven centuries later. An English monk, Eilmer of Malmesbury, is recorded as gliding a short distance in the 11th century and breaking both legs. None of these stories describes controlled flight.

Leonardo da Vinci studied the problem harder than anyone before him. His Codex on the Flight of Birds, written in 1505–06, holds more than 35,000 words and 500 sketches about bird flight, the nature of air and flying machines. In the early 1480s he had sketched an "aerial screw", a spiral meant to bore its way upwards, the first recorded design for vertical flight. There is no record that any of his flying machines flew. His aerial screw also had a flaw every helicopter designer would later meet: nothing stopped the machine itself from spinning the opposite way to its rotor.

1783: the year people left the ground

The breakthrough came from an unexpected direction: not wings, but buoyancy. Two brothers from a family of paper manufacturers, Joseph and Étienne Montgolfier, took their cue from something as ordinary as laundry drying over a fire and billowing upward. On 4 June 1783, at Annonay in southern France, they sent an unmanned balloon up in public; it flew for about ten minutes and came down some two kilometres away. On 19 September, at Versailles, a balloon carried a sheep, a duck and a rooster for about eight minutes, and all three landed safely. On 21 November, Jean-François Pilâtre de Rozier and the marquis d'Arlandes made that first crewed free flight over Paris. Ten days later, on 1 December, the physicist Jacques Charles rose about three kilometres in a balloon filled with hydrogen.

Joseph Montgolfier thought the smoke contained a special gas with a property he called "levity". He was wrong about why it worked, as pioneers often are. The real reason is the one that makes a steel ship float (see Everyday Physics): an object in a fluid is pushed up by a force equal to the weight of the fluid it displaces. Heat air from 15 °C to 100 °C and a cubic metre of it gets about 0.28 kg lighter, so every cubic metre of hot air can lift about 0.28 kg of balloon and passenger. Hydrogen does about four times better, roughly 1.1 kg per cubic metre, which is why Charles's balloon could be smaller and go higher.

A balloon floats; it doesn't fly. For the next 120 years the problem wasn't getting up. It was going where you meant to go.

Balloons went where the wind went. To steer against the breeze, you'd need a machine heavier than air that held itself up by moving.

Cayley, Lilienthal and the problem nobody saw

The man who first wrote down what an aeroplane is was an English baronet, George Cayley (1773–1857). In 1799 he engraved a small silver disc, about an inch across and now held by London's Science Museum, with a design for a fixed-wing flying machine. The idea it captured was radical: stop trying to copy a flapping bird. Use one surface, a fixed wing, to produce lift; a separate system to push the machine forward; and a third set of surfaces, like a tail, to control it. He is credited with naming the four forces every pilot still learns: weight, lift, drag and thrust.

Cayley flew a model glider with a modern layout in 1804 and published "On Aerial Navigation" in 1809–10. In 1853, at Brompton Dale in Yorkshire, a larger glider of his carried a person across a small valley. Popular accounts say it was his coachman, who promptly resigned. Historians aren't sure who it was: candidates include the coachman, a footman, a butler and his grandson. Wilbur Wright later said Cayley had taken the science of flight further than anyone before him.

What Cayley lacked was a light engine. Inventors tried steam anyway. On 9 October 1890, Clément Ader's Éole made an uncontrolled hop of about 50 metres in France; in 1894 Hiram Maxim's huge test rig lifted briefly off its rails and was wrecked; on 6 May 1896, Samuel Langley's unpiloted Aerodrome No. 5 made a sustained flight. Each proved power could make lift. None showed a person could fly a machine.

The flying man

The person who did most to show it was a German engineer, Otto Lilienthal. Between 1891 and 1896, with his brother Gustav, he made more than 2,000 flights in gliders of his own design, the longest about 250 metres. His 1889 book, Birdflight as the Basis of Aviation, was widely read. And he was photographed: at least 145 pictures show him in the air, hanging under great curved wings. For readers around the world, they were proof that a human being could glide like a bird.

On 9 August 1896 his glider stalled and fell about 15 metres. He died in Berlin the next day. The words on his gravestone, "Opfer müssen gebracht werden!" ("Sacrifices must be made!"), are often given as his last words, but the director of the Otto Lilienthal Museum doubts he ever said them.

Look at the last column. By the 1890s, almost everyone was working on lift and power. Very few were working on control: how a pilot keeps a machine level in gusty air, turns it, and gets out of trouble. Lilienthal's death came from exactly that gap. Two brothers in Dayton, Ohio, came to believe that control was the real problem.

Part 2 — 1903, and the long fight over who was first

Wilbur and Orville Wright ran a bicycle shop. On 30 May 1899 Wilbur wrote to the Smithsonian asking for whatever it had published on flight. What followed is best understood as three separate inventions.

1. Control in all three directions. A plane can tip its nose up and down (pitch), roll its wings (roll), and swing its nose left and right (yaw). The Wrights controlled roll by twisting, or "warping", the tips of their wings in opposite directions, pitch with a movable surface at the front, and yaw with a movable rudder at the back. Their 1902 glider made properly controlled turns on 9 October 1902. They filed their patent on 23 March 1903. This three-axis control is still how every aeroplane is flown.

2. Their own data. Their 1901 glider produced only about a third of the lift they had calculated. They traced much of the error to a long-accepted coefficient used in everyone's lift formulas, built a small wind tunnel in their shop in the autumn of 1901, and tested about 200 model wing shapes, 38 of them in detail. From then on they designed with their own measurements.

3. Propellers as spinning wings. Nobody had a theory of propellers. The brothers realised a propeller blade is a wing moving in a circle and designed theirs that way. With a 12-horsepower engine built with the help of their shop mechanic, Charles Taylor, the 1903 Flyer had two propellers behind the wings, driven by chains.

On 17 December 1903, at Kill Devil Hills near Kitty Hawk, North Carolina, Orville lifted off at 10:35 a.m. and flew 36 metres (120 feet) in 12 seconds. The brothers took turns. On the fourth flight that day, Wilbur flew 260 metres (852 feet) in 59 seconds. Soon afterwards a gust flipped the machine over on the ground. Severely damaged, it never flew again.

Then the Wrights went quiet. Back home in Ohio, at a cow pasture outside Dayton called Huffman Prairie, they added a weight-powered catapult for take-off from 7 September 1904. On 5 October 1905, Wilbur flew 39.4 kilometres in 38 minutes and 3 seconds, circling until the fuel ran out. Afraid that rivals would copy them, they refused to fly anywhere without a firm contract to sell their aircraft, and local newspapers largely ignored the flights they did make. Few people saw them. European aviators called them bluffers. When Wilbur finally flew in public at Le Mans in France on 8 August 1908, his smooth banking turns ended the argument overnight. The editor of the French magazine L'Aérophile wrote that the flights had "completely dissipated all doubts". On 17 September 1908 a propeller broke during a demonstration in Virginia; Orville was badly hurt and his passenger, Lieutenant Thomas Selfridge, became the first person killed in a powered aeroplane crash.

The credit disputes, presented honestly

Three rival stories are worth knowing, because each shows what the word "first" hides.

Alberto Santos-Dumont. In Paris, the Brazilian-born aviator flew his 14-bis 50 metres on 23 October 1906, winning a prize for the first flight of more than 25 metres, and on 12 November 1906 flew 220 metres in 22 seconds before officials of the Aéro-Club de France. His supporters say he deserves the title: his aircraft took off from wheels under its own power, in public, before official witnesses, while the Wrights flew in near-secrecy and from 1904 used a catapult. The case against: the 1903 Flyer used no catapult, just a launch rail and a headwind; by 1905 the Wrights were flying 39-kilometre controlled circles; and the 14-bis couldn't control roll until ailerons were added in November 1906.

Gustave Whitehead. Born in Bavaria in 1874, Whitehead emigrated to the United States in 1893 and was reported by the weekly Bridgeport Herald (in an article dated 18 August 1901) to have flown about half a mile at 50 feet on 14 August 1901. No photograph of the flight has ever been found, and a key witness later swore he saw none. In 2013 the editor of Jane's All the World's Aircraft backed Whitehead (Jane's later called it the editor's personal view), and Connecticut passed a law recognising him. His supporters point to something real: in 1948, to get the Wright Flyer back from London, the Smithsonian agreed with Orville's estate to recognise no earlier aircraft as first, on pain of losing the Flyer. That is an awkward thing for a museum to sign. But evidence, not contracts, settles history, and the Royal Aeronautical Society and the Smithsonian's historians find the evidence for Whitehead doesn't hold up. "No compelling evidence that Whitehead or anyone else flew before the Wright brothers," said the Smithsonian's Peter Jakab in 2005.

Langley and the Smithsonian. That 1948 agreement had a history. Samuel Langley, Secretary of the Smithsonian, saw his full-size piloted Aerodrome plunge into the Potomac in October and December 1903. In 1914 Glenn Curtiss, who was fighting the Wrights' patents, made short hops in a heavily modified Aerodrome, and the Smithsonian cited them as proof Langley's machine could fly, without disclosing the changes. Orville lent the Flyer to London's Science Museum in 1928 instead. The Smithsonian retracted in 1942, and the Flyer went on display in Washington on 17 December 1948, 45 years to the day after it flew.

The honest conclusion: "first" depends on the definition. For the first sustained, controlled, powered flight backed by solid evidence, the Wrights' claim is strong. For the first such flight before official judges, Santos-Dumont has a fair case. The Wrights' long patent war had a cost: when the United States entered the First World War in 1917, its forces flew French machines.

Part 3 — How a wing really makes lift

Ask most people how a wing works and you'll hear a version of this: the top of the wing is curved and longer, so air going over the top has further to travel; to meet the air from underneath at the back edge "at the same time", it must go faster; faster air has lower pressure (Bernoulli's principle); so the wing is sucked upward.

That explanation, called the equal transit time theory, is in many textbooks and is wrong. NASA's Glenn Research Center lists the problems plainly:

• There is no physical law that says the two streams of air must meet again at the back. Air that starts side by side at the front edge does not arrive together at the back.
• Wings with identical top and bottom surfaces make lift. So does a flat paper aeroplane. And planes can fly upside down, which the theory can't explain.
• If you work out the speed the theory predicts, it is far too small to produce the lift real wings make.

You can see the failure on film. In 2012, Holger Babinsky of Cambridge University released wind-tunnel footage of puffs of smoke flowing past a wing. Pause it and the smoke over the top has already passed the back edge while the smoke underneath is still on its way. The air over the top doesn't just keep pace; it arrives first, much faster than equal transit predicts.

"It is the turning that's important, not the distance." NASA Glenn Research Center, on how wings make lift

So what is going on? A wing meeting the air at a slight upward angle, its angle of attack, turns the flow. Air streaming past is bent downward behind the wing. Newton's third law then does the rest: if the wing pushes air down, the air pushes the wing up. Turning a flow also requires pressure differences, low above the wing and higher below it, and those are what you'd measure with gauges on the wing's surfaces. The "Newton" story (downwash) and the "Bernoulli" story (pressure and speed) are not rivals: they describe the same flow, and you need both for the whole picture. Physicists still argue over how best to explain, simply, why the flow follows the curved upper surface; the honest short answer is that it does, as long as the angle isn't too steep.

The lift equation

Engineers sum all of this up in one line:

Lift = ½ × ρ × v² × S × C_L

Here ρ (rho) is the density of the air, v is the speed through it, S is the wing's area and C_L, the lift coefficient, captures the wing's shape and its angle of attack. Each term tells you something practical:

• Speed is squared. Double your speed and you get four times the lift. This is why planes need runways: they must reach a speed at which lift equals weight.
• Density matters as much as speed. At 11,000 metres, air is about 3.4 times thinner than at sea level, so at the same angle a plane must fly about 1.8 times faster to stay up.
• Angle helps, until it doesn't. Up to a point, lift rises roughly in step with angle of attack. Beyond about 10–15° for a typical wing, the air can no longer follow the upper surface. It breaks away, lift collapses and drag soars. That is a stall, and it is what killed Lilienthal.
• Flaps change the shape. At take-off and landing, airliners slide out flaps and slats that make the wing bigger and more curved, raising its maximum lift so they can fly slowly without stalling.

Use the explorer below to feel how these trade off.

Three things to notice. At 100 km/h the Cessna's wing makes only about 60% of its weight; it needs about 128 km/h at that angle. Cruising at 837 km/h at 11,000 metres, lift almost exactly matches a fully loaded 737's weight, but at the same speed and angle at sea level the wing would make 3.4 times the weight, so the pilot would lower the nose to about 0°. And past about 14° the model gives up, because the real airflow does too.

The model Cessna stalls at about 101 km/h with flaps up; the real one's book stall speed with flaps down is 47 knots, about 87 km/h. That difference is what flaps are for.

Part 4 — Wars, jets and the shrinking planet

Aeroplanes went to war almost as soon as they could fly. When the First World War began in August 1914, Germany's army had about 180 usable aircraft, used first for reconnaissance: finding troops, photographing trenches, spotting artillery fire. Pilots soon carried pistols, then machine guns. The problem of shooting through your own propeller was solved by a synchronising gear that fired only when no blade was in the way; Fokker's E.I fighter, which entered service in July 1915, gave Germany a period of air superiority the Allies called the "Fokker Scourge". Bombing from the air began too: German airship raids on England, from 1915 to 1918, killed 557 people.

The Second World War turned aircraft into an industry on an unprecedented scale. The United States alone built more than 300,000 military aircraft between 1939 and 1945 (sources differ on the exact total). The war also pushed through the invention that defines modern flight.

The jet engine: invented twice

A piston engine turning a propeller loses efficiency as a plane flies faster and higher. The jet engine works on a different idea, which mechanics sum up as suck, squeeze, bang, blow: a compressor sucks in air and squeezes it, fuel burns in it, and the hot gas rushes out the back, spinning a turbine on its way that drives the compressor. The plane is pushed forward by throwing a stream of gas backward, Newton's third law again.

It was invented twice, independently. Frank Whittle, a young Royal Air Force officer, filed his patent on 16 January 1930. In January 1935 it lapsed because the Air Ministry refused to pay the £5 renewal fee and he couldn't afford it. With private backers he formed Power Jets in 1936, and his first engine ran on 12 April 1937. Meanwhile in Germany, Hans von Ohain, initially unaware of Whittle's work, had his first engine running by September 1937, working with the aircraft maker Heinkel. On 27 August 1939, days before the war began, the Heinkel He 178 made the world's first flight powered by a jet. Britain's first jet, the Gloster E.28/39, flew on 15 May 1941. Germany's Me 262 became the first jet fighter in 1944.

Modern airliners use a refinement, the turbofan: a big fan at the front, driven by the jet core, pushes most of the air around the core rather than through it. Moving a lot of air gently turns out to be more efficient than moving a little air violently, which is why airliner engines have grown so fat.

Speed records fell. On 14 October 1947, Chuck Yeager flew the rocket-powered Bell X-1 at Mach 1.06, faster than sound, after being dropped from the bomb bay of a B-29.

The jet age, and the price of a lesson

The first jet airliner, Britain's de Havilland Comet, carried fare-paying passengers from London to Johannesburg on 2 May 1952. In 1954 two Comets broke up in flight over the Mediterranean, killing 35 people on 10 January and 21 on 8 April. Engineers pressurised a whole fuselage in a water tank again and again; it burst after 3,057 simulated flights, far fewer than expected. Metal fatigue had started from cracks at the corners of square cut-outs for radio antennas in the roof of the fuselage (not, as is often said, at the passenger windows). The lesson shaped every airliner since: stricter pressure testing and "fail-safe" structures in which a single crack can't bring the aircraft down.

The Boeing 747, the first wide-body and the original "jumbo jet", first flew on 9 February 1969 and entered service with Pan Am between New York and London on 22 January 1970. Pan Am had asked for a plane that would carry each passenger a given distance for about 30% less. In three classes it seated about 366 people. Boeing built 1,574 of them; the last was delivered on 31 January 2023.

Concorde went the other way: speed, not seats. The Anglo-French airliner first flew on 2 March 1969 and entered service on 21 January 1976, cruising at about twice the speed of sound. Its sonic boom kept it off routes over land, the 1973 oil crisis made its fuel thirst frightening, its costs ran more than six times over estimates, and airlines cancelled their options. Only 20 were built. On 25 July 2000 an Air France Concorde ran over a metal strip dropped by another airliner; a burst tyre ruptured a fuel tank, and the burning aircraft crashed into a hotel, killing 113 people. It was Concorde's only fatal accident. The last commercial Concorde flight was on 24 October 2003.

The jumbo jet and Concorde both flew in 1969. The future belonged to the one that made flying cheaper, not faster.

Part 5 — Why flying became so safe

Here is a statistic worth carrying around in your head. In 1970, airliners worldwide had about 6.5 fatal accidents for every million flights, according to Aviation Safety Network data compiled by Our World in Data. In 2017 the figure was about 0.28, more than twenty times lower. It bounces around from year to year (it was 0.48 in 2022), but the long-term fall is enormous, and it happened while the number of air passengers grew about fourteen-fold, from 310 million in 1970 to 4.5 billion in 2019.

Note that the IATA rows count all accidents, including non-fatal ones, so they are higher than the fatal-accident rows. Note too that 2025 was clearly worse than 2024 on deaths: averages hide bad years, and every one of those 394 deaths was a person. But even in a bad year, one accident of any kind for roughly every 760,000 flights is a remarkable number for a machine that cruises at 800 km/h, ten kilometres up.

Put the 2024 figure in personal terms. One accident per 880,000 flights means that if you boarded a flight every single day, you would expect to be on board for an accident of any kind, most of them non-fatal, about once in 2,400 years. In the Our World in Data series, which runs from 1970 to 2022, the safest year by fatal accidents per flight was 2017. And 2025, for all its losses, was the first year in IATA's records with no accident in which a crew lost control of an aircraft in flight, the category that includes stalls like the one that killed Lilienthal.

How did flying get this safe? Not through any single invention, but through a habit of learning from every failure.

• Engineering lessons. The Comet taught designers to expect fatigue cracks and build structures that survive them.
• Human lessons. On 27 March 1977, two Boeing 747s collided on a fog-covered runway at Tenerife after a garbled radio exchange led one captain to start his take-off without clearance; 583 people died, still the deadliest accident in aviation. Afterwards, authorities worldwide adopted standard radio phrases. On 28 December 1978, a United Airlines crew near Portland, Oregon, ran out of fuel while busy with a landing-gear problem. These accidents led to crew resource management, a term coined at NASA in 1979: training that teaches junior crew to speak up and treats the captain as fallible. United began the first full programme in 1981.
• Technology that watches the pilots' backs. Terrain warning systems shout when an aircraft is too close to the ground. By February 2008, more than 40,000 aircraft carried the enhanced version and had flown over 800 million hours without a single accident of flying a working aircraft into terrain.
• Investigation in the open. Each serious accident is investigated, the findings are published, and they change training, procedures and design for airlines everywhere, as Tenerife and the Comet show.

The lesson runs beyond aviation: a system becomes very safe when every failure is studied in the open and every finding changes the rules for everyone.

Part 6 — Rotors and robots: helicopters and drones

A helicopter's rotor blades are wings too, just whirling in a circle so that air flows over them even when the aircraft is standing still. The hard part, as da Vinci's aerial screw showed, is everything else. Spin a rotor one way and the body tries to spin the other way. Fly forward and the blade moving forward meets faster air than the blade moving backward, so one side of the rotor lifts more than the other. Solving those problems took decades.

On 13 November 1907, Paul Cornu's twin-rotor machine in France lifted him about 30 centimetres for 20 seconds. In the early 1920s the Spanish engineer Juan de la Cierva invented the autogiro, whose unpowered rotor spins in the airflow like a sycamore seed, the first practical rotorcraft. Germany's Focke-Wulf Fw 61, which first flew in June 1936, was the first fully controllable helicopter. The layout most helicopters still use, one big main rotor and a small tail rotor pushing sideways to stop the body spinning, came from Igor Sikorsky's VS-300, which he developed in the United States from 1939 to 1943. His R-4 of 1942 was the first helicopter built in full-scale production.

A pilot flies a helicopter with three controls. The collective changes the angle of all the main blades together, for more or less lift. The cyclic changes each blade's angle as it goes round, tilting the rotor disc so the aircraft moves forward, backward or sideways. Pedals vary the tail rotor's push to point the nose. In the Korean War, helicopter evacuation cut the time to get a wounded soldier to hospital from about eight hours in the Second World War to about three.

Drones went from target practice to every pocket and every front line in less than a century.

Drones: from target practice to every pocket and every front line

Pilotless aircraft are older than you might think. In July 1849, Austrians besieging Venice launched about 200 bomb-carrying balloons; most missed when the wind changed. In 1918 the United States flew the Kettering Bug, an "aerial torpedo" that was not ready before the war ended. In 1935 Britain produced the Queen Bee, a radio-controlled target plane, and its name is said to be where the word "drone" comes from. During the Second World War, Reginald Denny's company built nearly 15,000 radio-controlled target drones for the US Army.

Military drones became central in the following decades. In 1982 Israel coordinated drones with piloted aircraft in a lopsided victory over the Syrian air force. A US drone strike in November 2002 was an early example of armed drones used for targeted killing, a practice that remains legally and morally contested. In the war that began with Russia's 2022 invasion of Ukraine, both sides use drones in vast numbers, from cheap racing-style quadcopters to long-range attack drones; Ukraine created a separate branch of its armed forces for unmanned systems in 2024, and Russia followed in 2025.

The civilian drone story is mostly about one company. DJI, the Chinese company founded in 2006 by Frank Wang, launched its Phantom quadcopter in 2013 and by mid-2024 held more than 90% of the world's consumer drone market. That success became a political issue: in December 2025 the US Federal Communications Commission banned the import, marketing and sale of new DJI drone models in the United States. Quadcopters themselves are a triumph of electronics more than aerodynamics: four fixed propellers, cheap motion sensors and a computer adjusting each motor's speed many times a second. To fly forward, the computer speeds up the rear rotors a little, tilting the whole drone so that part of its thrust points forward, a trick no human could manage by hand at that speed.

Part 7 — Aviation in 2026, and the open questions

IATA counted just under five billion passengers and more than 38 million flights in 2025. Growth has slowed: on 30 October 2026 IATA reported that passenger demand in September 2026 was 0.5% lower than a year before. The big questions now are about carbon and what might replace, or join, the jet.

How much does flying warm the planet?

In 2019 aviation produced about 2.5% of the world's carbon dioxide emissions from fossil fuels and land use, according to Our World in Data. That understates its climate effect, because planes also warm the planet through contrails, trails of ice crystals that can spread into heat-trapping high cloud. Counting those effects, a major 2021 study by David Lee and colleagues estimated that aviation accounted for about 3.5% of the human-caused warming push on the climate, with non-CO₂ effects making up about two-thirds of aviation's share. (For the physics of why CO₂ warms the planet, see The Physics of the Greenhouse Effect.) The non-CO₂ estimates are more uncertain than the CO₂ ones.

Sustainable aviation fuel: the main bet, still tiny

Airlines have committed to net-zero carbon emissions by 2050, and their trade body, IATA, says sustainable aviation fuel (SAF) would deliver about 65% of the cuts. SAF is jet fuel made from used cooking oil, farm and forest waste or, eventually, captured carbon and renewable electricity, blended with ordinary kerosene in today's engines. The most common type, made from waste fats and oils, cuts lifecycle emissions by about 80%, says IATA, because its carbon was recently taken from the air.

Source: IATA SAF fact sheet, June 2026.

The growth is real, but the base is tiny and SAF costs several times more than fossil jet fuel. The European Union's ReFuelEU law requires at least 2% SAF from 2025, rising to 70% by 2050, and the UK has a similar mandate. Supporters say only guaranteed demand will get factories built. IATA, for the airlines, argues that the mandates have so far failed to speed up production and have raised costs, and that governments should support production directly. The hard part is supply. Today's SAF comes mostly from waste fats and oils, of which there is only so much, so most future SAF would have to come from newer processes that are not yet running at scale.

Electric planes: the battery problem

Why not use batteries, as cars increasingly do? Because weight is everything in the air. Jet fuel holds about 43 megajoules of energy per kilogram; lithium-ion batteries hold between about 0.36 and 0.9. Electric motors are far more efficient than jet engines, which narrows the gap, but nowhere near enough for a large airliner on a long route. And a battery doesn't get lighter as it empties.

So electric flight has started small. The first fully electric aircraft to earn a type certificate was the two-seat Pipistrel Velis Electro, certified by the European Union Aviation Safety Agency on 10 June 2020. It flies for about 50 minutes plus a reserve and is used mainly for pilot training.

Air taxis: where they really stand

The most talked-about new aircraft are eVTOLs (electric vertical take-off and landing aircraft) with several battery-powered rotors, pitched as quiet air taxis. Their status as of late October 2026:

Boosters argue that eVTOLs will be quieter and cleaner than helicopters, cheap per trip once built in numbers, and a way to use the sky above jammed cities. Critics point to short ranges, heavy batteries, the cost of building landing sites, the need to prove a new kind of aircraft as safe as airliners, and the doubt that air taxis will ever be more than a premium service. The record so far supports both: real certification progress in China and the US, and European pioneers that ran out of money first.

The air taxi has been "two years away" for a decade. In 2026 it is, for the first time, carrying paying passengers in at least one country.

Open questions and trade-offs

Can aviation reach net zero by 2050? The airlines' plan rests on SAF growing from under 1% of fuel to most of it in 24 years, plus better aircraft and offsets. Optimists point to how fast solar power scaled once costs fell; sceptics point to feedstock limits and argue demand growth must slow. Neither side can prove its case yet.

What about contrails? If non-CO₂ effects really are about two-thirds of aviation's warming, avoiding the worst contrails could be one of the cheapest climate measures there is, but it means flying slightly longer routes and burning a little more fuel. Weighing a certain small extra CO₂ against an uncertain contrail benefit is an active scientific argument.

How much should we automate? In September 2026 Joby flew a converted Cessna Caravan 3,199 miles across the United States with its safety pilot making no control inputs, and pilotless air taxis already carry passengers in China. Technology that watches over pilots has made flying safer; taking the pilot out entirely is a different step, and it will test how much trust passengers and regulators give a machine.

Who gets the sky? Drones, air taxis and helicopters will increasingly share low airspace, bringing benefits along with noise, privacy and safety worries. The rules are being written now.

And one lesson from 1903 still applies. The Wrights succeeded not through money or a big engine but by finding the real problem, control, and testing everything themselves. That is still how flying gets better.