Every device had to work in deep space with no repair option.
At a distance four times beyond the Moon, the James Webb Space Telescope carried into the void a list of 344 ways it could silently fail — each one irreversible, none repairable. What unfolded over ten days in January 2022 was not merely an engineering sequence but a kind of high-stakes origami: a civilization's most complex scientific instrument trusting itself, piece by piece, to the indifferent physics of deep space. That it succeeded — retiring 295 of those 344 catalogued risks — is less a story of technology than of accumulated human patience, humility before uncertainty, and the quiet courage of committing to something that cannot be undone.
- With 344 catalogued single-point failures and no possibility of repair 1.5 million kilometres from Earth, Webb launched as perhaps the most consequential all-or-nothing gamble in the history of scientific instrumentation.
- The sunshield alone — a tennis-court-sized, five-layer kapton origami — required 107 individual release devices to fire in sequence, each one a silent veto over the entire mission if it refused.
- Tensioning the sunshield's layers in microgravity introduced interaction risks between cables, pulleys, motors, and membranes that Earth-based testing could only approximate, leaving engineers to trust years of iterative redesign over lived certainty.
- On 4 January 2022, all 107 devices fired, all five layers tensioned to specification, and a single announcement retired between 70 and 75 per cent of the mission's accumulated risk in one breath.
- When full deployment was confirmed, 295 of 344 single-point failures were officially retired — the remaining 49, standard propulsion systems, will accompany the telescope for the rest of its operational life.
The James Webb Space Telescope orbits the Sun-Earth L2 point, 1.5 million kilometres from Earth, drawing roughly one kilowatt of electrical power — less than a household kettle. Its four science instruments, communications, propulsion, and thermal systems all run on what it takes to boil water for breakfast. That economy of power is a design philosophy, not a compromise: most of the telescope's cooling happens passively through the sunshield, allowing its instruments to operate near absolute zero without power-hungry cryocoolers.
None of that elegance would have mattered if the deployment had failed. Webb launched carrying 344 officially catalogued single-point failures — components or procedural steps where a single malfunction would have ended the mission permanently. About 80 per cent of those risks were embedded in the deployment sequence itself. NASA and prime contractor Northrop Grumman maintained a meticulous inventory: 140 release mechanisms, 70 hinge assemblies, 8 deployment motors, 400 pulleys, roughly a quarter mile of cable, and 178 release devices on the primary mirror alone. Every one had to function in deep space with no option for repair.
The sunshield was the most unforgiving element. Approximately 21 by 14 metres — the size of a tennis court — it consists of five separated layers of kapton film, folded origami-style for launch and designed to unfurl, separate, and tension itself over roughly a week in space. The 107 non-explosive actuators holding the membranes in place during launch had been reduced from 109 through years of redesign. Tensioning the layers proved the hardest step to simulate on Earth, because the interactions between cables, pulleys, motors, and membrane behave differently in microgravity than in gravity.
All 107 devices fired. All five layers reached their final tension. The sunshield completed its deployment on 4 January 2022, ten days after launch, retiring between 70 and 75 per cent of the original 344 single-point failures in a single sequence. When full deployment was confirmed, 295 of those 344 risks had been officially retired. The remaining 49 — primarily propulsion systems — will remain on the risk list for the mission's duration. The 155 motors fine-tuning the 18 hexagonal mirror segments were each tested individually afterward. Every one functioned. The telescope, running on a kilowatt of power at the edge of the Earth-Moon system, began returning infrared images no other instrument has been able to produce.
The James Webb Space Telescope sits in a halo orbit around the Sun-Earth L2 point, roughly 1.5 million kilometres from Earth, drawing about one kilowatt of electrical power. To put that in perspective: a household kettle uses more. The solar array generates close to two kilowatts to account for degradation over the mission's lifetime, but the entire spacecraft—its four science instruments, communications systems, propulsion, and thermal management—operates on roughly what it takes to boil water for breakfast.
None of this engineering elegance would have mattered if the deployment had failed. According to Mike Menzel, Webb's lead mission systems engineer at NASA's Goddard Space Flight Center, the observatory launched carrying 344 single-point failures on its official list. In engineering terms, a single-point failure is any component or procedural step where one malfunction ends the mission. About 80 per cent of those 344 items were embedded in the deployment sequence that had to unfold perfectly in the vacuum of space.
The number was not casual speculation. NASA and Northrop Grumman, the prime contractor, maintained a meticulous inventory of every mechanism that could not afford to break: 140 release mechanisms, 70 hinge assemblies, eight deployment motors, 400 pulleys, and roughly a quarter mile of cable strung through the structure. The primary mirror alone carried 178 release devices. The five-layer sunshield, which shields the instruments from the sun's heat, required 107 non-explosive actuators to hold the kapton membranes in place during launch. Scott Willoughby, Northrop Grumman's Webb programme manager, noted that the team had actually reduced that number from 109 through years of iterative redesign. Every single device had to function in deep space with no realistic option for repair or replacement.
The sunshield emerged as the deployment's most unforgiving component. Roughly 21 metres by 14 metres—about the size of a tennis court—it consists of five separated layers of kapton film coated with aluminium and silicon. The first layer measures 50 microns thick; the others are 25 microns each. The entire structure folds origami-style for launch and then unfolds, separates, and tensions itself over approximately a week in space. Tensioning proved to be the step that introduced the greatest interaction risk between cables, pulleys, motors, and membrane. James Cooper, NASA's Webb sunshield manager, explained at the time that tensioning was the hardest part of the deployment to simulate on Earth, because the complex interactions between structures and mechanisms behave differently in Earth's gravity than they do in microgravity.
All 107 release devices fired. All five layers achieved their final tension. The sunshield reached full configuration on 4 January 2022, ten days after launch. Bill Ochs, then JWST project manager, announced that completing the sunshield deployment had retired between 70 and 75 per cent of the original 344 single-point failures.
The reason the deployment had to work perfectly on the first attempt comes down to geography and physics. L2 sits 1.5 million kilometres away—roughly four times the Earth-Moon distance. Unlike the Hubble Space Telescope, which orbits close enough to Earth that the Space Shuttle visited it five times before retirement in 2011, Webb has no servicing option. The location was chosen precisely because it allows the telescope to observe the universe without Earth's heat interfering with its infrared instruments.
The one-kilowatt power budget reflects the same constraint. Webb's instruments operate at extreme cold: the MIRI detector runs at around 7 kelvin, while the rest of the optical assembly hovers near 40 kelvin. Most of this cooling happens passively through the sunshield rather than through power-hungry mechanical cryocoolers. Passive cooling draws no electricity. The result is a spacecraft that can observe, store, and transmit science data on roughly the electrical draw of a kitchen appliance.
When deployment was complete, NASA and the Space Telescope Science Institute reported that 295 of the original 344 single-point failures had been retired. The remaining 49 are standard spacecraft systems—primarily the propulsion system—that will remain on the risk list for the mission's duration. The 155 small motors mounted on the backs of the 18 hexagonal primary mirror segments, which fine-tune the optics, were each tested individually after deployment. Every one functioned. The telescope, running on a kilowatt of power 1.5 million kilometres from Earth, began returning infrared images no other instrument has been able to produce.
Notable Quotes
Tensioning was the hardest part of the deployment to test on the ground, because the complex interactions between structures, mechanisms, cables, and membranes do not behave the same way in 1 g as they do in deep space.— James Cooper, NASA's Webb sunshield manager
Completion of the sunshield retired between 70 and 75 per cent of the 344 single-point failures on the original list.— Bill Ochs, then JWST project manager
The Hearth Conversation Another angle on the story
Why does a telescope that complex need to run on so little power?
Because it's so far away. At L2, you can't send a repair mission. So the design had to be elegant—passive cooling through the sunshield instead of power-hungry refrigeration. It's a constraint that forced brilliance.
And the 344 single-point failures—that sounds terrifying. How do you even launch something with that many ways to fail?
You don't have a choice. The sunshield alone has 107 release devices. The mirror has 178. You catalogue every one, you test obsessively, and you accept that if any single device fails, the mission is over. There's no backup plan at L2.
So when the sunshield deployed successfully, that was the moment the mission actually became viable?
Exactly. Completing the sunshield retired 70 to 75 per cent of the failure risks. Once those five layers tensioned properly in microgravity, the hardest part was done. The rest was almost routine by comparison.
Why was tensioning so hard to predict on Earth?
Gravity changes everything. On the ground, you can't test how cables and membranes interact in weightlessness. The physics is fundamentally different. You can simulate it, but you can't truly know until you're in space.
And all 155 mirror motors worked on the first try?
Yes. Every single one. That's the part that still feels like luck, even though it was careful engineering. But at that distance, luck is all you have.