The James Webb Space Telescope is often described by what it can see: galaxies forming near the dawn of time, planets around other stars, dust clouds where stars are being born. But the more astonishing part may be how little room there was for error before any of that science could begin.
Webb had to leave Earth folded inside a rocket, unfold itself in space, fly to a gravitational parking region roughly a million miles away and then operate on about the electrical power of a small kitchen appliance. NASA’s own fact sheet puts Webb in orbit around the Sun-Earth L2 point, 1.5 million kilometres from Earth. That is far beyond the kind of astronaut servicing mission that rescued Hubble.
The other number is even more unsettling. Northrop Grumman, Webb’s industry team leader, says the finished observatory carried 344 single-point failures, including 295 associated with deployable systems. A single-point failure is a part, action or mechanism whose failure could not simply be worked around. If it did not do its job, the mission could be diminished, crippled or lost.
That is why Webb’s first weeks in space became one of the most watched engineering sequences NASA has ever flown. Before the telescope could become an observatory, it had to become itself.
Webb was too large to launch open
Webb’s scientific power comes from scale and cold. Its primary mirror is 6.5 metres wide, made of 18 gold-coated beryllium segments. Its five-layer sunshield is roughly the size of a tennis court. The cold side of the observatory has to stay extremely cold so its infrared instruments can detect faint heat from distant objects rather than the telescope’s own warmth.
None of that fit inside a rocket in one open piece. NASA’s deployment page says Webb’s operational configuration is about 70 feet long, 47 feet wide and 28 feet tall, far too large for any available launch fairing. It therefore had to be folded compactly inside the Ariane 5 rocket and then unfurled, extended and expanded after launch.
The basic idea sounds elegant. The execution was unforgiving. NASA says the deployment process included more than 50 major deployment steps and 178 release mechanisms that all had to work properly. The sunshield alone required membranes, cables, pulleys, motor-driven systems, booms, release devices and tensioning steps to behave in the right order.
On Earth, a jammed mechanism can sometimes be reached by a technician. At L2, it cannot. Webb’s deployment therefore had to be tested, rehearsed and analysed before launch with a level of obsession that matched the risk. Engineers had to know not only how each part worked, but how it behaved after vibration, cold, folding, storage, launch loads and the transition to vacuum.
The 344 number came from real hardware
NASA often described Webb before launch as having “over 300” single-point-failure items. Northrop Grumman later gave the more precise figure of 344. The company’s account says 295 of them were tied to deployable systems, many in the sunshield. Its description of Webb’s testing effort says there were more than 344 possible single points of failure in the system.
Those were not abstract spreadsheet anxieties. A membrane release device that failed to release could leave part of the sunshield trapped. A pulley or cable problem could keep a layer from reaching the right shape. A mirror wing that failed to latch could prevent the optical system from becoming a stable telescope. A secondary mirror support structure that did not deploy would leave the primary mirror with nowhere useful to send light.
The phrase “single point of failure” can make the mission sound reckless, but the opposite is closer to the truth. Webb had single points of failure because it was trying to do something that could not be made simple: launch a giant cryogenic infrared telescope compactly, then deploy it remotely. Engineers spent years reducing risk where possible and proving the remaining risk could be survived.
Northrop’s account of the sunshield notes that 107 membrane release devices were themselves single-point failures. That is a breathtaking sentence because the sunshield was not decorative. It was the thermal foundation of the mission. Without its five layers properly deployed and separated, Webb’s instruments would not have reached the cold conditions required for infrared astronomy.
One kilowatt for a flagship observatory
Webb’s power figure is the part that feels least intuitive. NASA described the telescope’s solar array as its “powerhouse,” but the total operational demand is small by household standards. NASA wrote in 2020 that one kilowatt is enough to power the largest and most technically advanced telescope ever built, and that Webb would use only 1 kilowatt of power while its solar array could generate nearly twice that amount.
That makes the kettle comparison fair, with a caveat. Household kettle power varies by country and model, but many electric kettles draw around 1,500 watts. A common retail guide from Home Depot notes that most kettles use about 1,500 watts. Webb’s ordinary operating draw is lower than that.
This does not mean Webb is underpowered. It means the observatory is engineered around efficiency. The solar array supplies the spacecraft bus, science instruments, communications system, propulsion system, heaters and electronics. The telescope is not firing a giant engine or heating a large volume of air. Its hardest thermal task is almost the opposite: keeping the cold side cold.
NASA’s telescope overview says the solar array provides about 2,000 watts of electrical power for the life of the mission, with margin above the roughly one-kilowatt use case. The array had to deploy quickly after launch because it stopped the drain on Webb’s internal battery and began powering the observatory’s systems directly.
The deployment succeeded step by step
Webb launched on 25 December 2021. The first deployment came quickly: the solar array extended less than an hour after launch. NASA’s post-launch deployment timeline placed the high-gain antenna deployment at about two hours, followed by trajectory correction manoeuvres, sunshield pallet deployments, tower extension, sunshield cover release, boom extension, membrane tensioning, secondary mirror deployment and primary mirror wing deployments.
NASA’s deployment summary says the full transformation unfolded over the first two weeks after launch. The first week was dominated by the sunshield. The second completed the telescope structure, including the secondary mirror tripod and the primary mirror wings. After that came cooldown, mirror alignment, instrument commissioning and insertion into the halo orbit around L2.
The sequence worked. That sentence is short because the work behind it was not. Webb did not simply open like a flower. It released launch locks, moved structures, extended booms, tensioned membranes and latched mirrors through a choreography that mission controllers monitored through telemetry from Earth.
NASA later wrote that Webb’s deployment was completed with remarkable smoothness and that the final results exceeded requirements and expectations. The spacecraft had survived the phase where most of the mission’s mechanical risk was concentrated.
Why L2 made repair almost impossible
Hubble’s early optical flaw became a famous story of rescue because Hubble orbited low enough for astronauts to reach it. Webb was never designed that way. Its observing site near Sun-Earth L2 gives it a stable thermal and viewing environment, keeping the Sun, Earth and Moon on the same side of the observatory’s sunshield. That geometry is central to its science.
The tradeoff is distance. A million miles is not merely high orbit. It is roughly four times farther than the Moon. Northrop Grumman described Webb as being stationed beyond the kind of manned rescue mission that saved Hubble, where “almost perfect” would not be good enough. It had to work in place.
That is why the 344 single-point-failure count is so arresting. Webb’s engineering story is not only that many things could have gone wrong. It is that the project reached launch with each of those risks understood deeply enough to proceed, then watched the observatory perform the necessary sequence in space.
The telescope now sits in its distant orbit, using less power than many kettles to run one of humanity’s most sensitive scientific instruments. That contrast is the real wonder of Webb. Not just a golden mirror seeing old light, but a fragile, folded machine that survived 344 ways to fail before it ever took its first science image.
