A truck wedged under a bridge, its roof peeled back like a sardine can, is such a familiar sight that the Australian city of Melbourne turned one bridge into a folk legend. The Montague Street rail bridge has claimed so many trucks that someone built a website tracking how many days since it was last hit. Bridge strikes are not exotic. They are one of the most repetitive failures in road transport.
Which brings us to a stretch of Central New York in the spring of 2026, and a load that never should have been on the road at the height it was.
What happened over 60 miles of the Thruway
On March 18, 2026, a tractor-trailer hauling an over-height excavator on a flatbed struck multiple overpasses as it traveled west on the Thruway.
The driver, 38-year-old Oleksander Tarakanovskyi of North Carolina, was apparently unaware that his load had struck the bridges. After the Thruway Authority alerted State Police, troopers found the truck and stopped it at the Mattydale tandem lot in the town of Salina.
Tarakanovskyi was charged with reckless endangerment in the second degree and issued more than 20 traffic tickets, including violations of federal commercial motor vehicle safety regulations. He was released with an appearance ticket returnable to Salina Town Court on April 8.
The public record does not explain how Tarakanovskyi failed to notice the impacts. But the larger bridge-strike problem often starts before a truck reaches the bridge: with the vehicle’s height and route.
A Network Rail survey in the UK, part of its “Wise Up, Size Up” campaign, found that 43 per cent of lorry drivers admitted they did not measure their vehicle before setting off, while 52 per cent said they did not take low bridges into account. That’s UK data, not a direct explanation for this New York incident, but it shows why height checks and route planning are central to bridge-strike prevention.
Permits, approved routes, escorts and height poles are all meant to move the moment of discovery to before the trip, rather than after a bridge is hit.
Take the King Street Bridge over the Hutchinson River Parkway in Westchester County. It was
The wider problem has been documented for years. A 2013 City College of New York study found that bridges in the state were experiencing roughly 200 strikes annually at the time, with many concentrated on a much smaller group of repeatedly hit bridges.
A solvable problem that stays unsolved
At a 2018 meeting about bridge strikes on the Hutchinson River Parkway, Westchester County Executive George Latimer captured the exasperation. “This is a solvable problem,” he said, adding, “If government can’t solve a problem like this, then government can’t do anything.”
The engineering measures are familiar: know the vehicle’s height, plan a legal route, use permits and escorts where required, and install detection systems where repeated strikes justify them. The hard part is making those checks work consistently across a large road network.
The 2026 Thruway incident stands out because one over-height load allegedly produced a 60-mile trail of damage before police stopped it. But the underlying failure was not new. Bridge strikes are repetitive precisely because the safeguards depend on measurements, route choices and warnings being acted on before the truck reaches the bridge.
A mother rinsed her baby’s dirty nappy in a bucket and tipped the water into the cesspool outside her front door. Within ten days, more than 500 of her neighbours were dead. If Henry Whitehead’s later reconstruction was right, hers was where it started.
The baby was five months old and lived at 40 Broad Street, Soho, in a district of tenements chopped into cheap rented rooms. Her death, on 2 September 1854, went into the records as diarrhoea rather than cholera, so she never appeared in the tally that made the street famous, as Gavi’s account of the outbreak notes.
Ten minutes’ walk away in Sackville Street lived a physician named John Snow, who had spent five years arguing that cholera travelled in water and getting nowhere. His 1849 essay on the subject was largely ignored by the medical profession. Respectable opinion held that epidemic disease rose off rotting matter as a poisonous vapour, a theory called miasma. Under miasma, the smell was the disease.
The map was made of black bars
Snow got the addresses of the dead from the General Register Office and walked them. He knocked on doors, asked survivors where they drew their water and worked out that the deaths clustered overwhelmingly around one public pump on Broad Street.
Months later he turned that investigation into the map that became famous, marking each known death as a short black bar at the address where it happened. By December he had tracked down 616 deaths, though only 574 could be placed on the map itself, the rest belonging to people who had died away from the district with no address on record. He published it in the second edition of On the Mode of Communication of Cholera in 1855.
He had no germ, no culture and no microscope image worth anything, only addresses.
The strongest evidence came from people who survived
Seventy men worked at the brewery on Broad Street and not one of them died. They drank a daily beer allowance, used a private well and never touched the public pump, per the Science History Institute’s account of the investigation. A workhouse squarely inside the death zone held over 500 people, lost only a handful and had its own supply.
Then there was the widow. A woman in Hampstead, miles to the north, died on 2 September, her visiting niece the day after. Neither had set foot in Soho for years. Snow went up to interview her son and learned she had once lived on Broad Street, liked the taste of the water and kept a servant carting bottles of it north. Her last delivery arrived on 31 August, the night the outbreak began.
Geography was the wrong variable and plumbing was the right one.
The parish agreed and the government did not
On the evening of 7 September, Snow put his case to the Board of Guardians of St James parish. They were sceptical of the theory and impressed by the arithmetic, and the handle was unscrewed the following morning.
Deaths were already falling by then, as Snow himself acknowledged. Thousands had fled the neighbourhood, and the contamination that triggered the outbreak had probably washed through days before.
The national verdict was blunter. England’s General Board of Health reviewed Snow’s theory through its Committee for Scientific Inquiries and rejected it outright: “After careful inquiry we see no reason to adopt this belief.” That committee, as Nigel Paneth and colleagues documented in the American Journal of Public Health, spent roughly a third of its 352-page scientific appendix on barometric pressure, wind direction, cloud cover and ozone levels. One of its own inspectors had handed Snow the Hampstead evidence.
The curate who found the first case
Henry Whitehead, assistant curate of St Luke’s on Berwick Street, had spent the outbreak climbing those stairs with food and last rites. He thought Snow was wrong, and joined the parish inquiry committee partly to clear the pump’s name. Snow gave him a copy of the monograph and he read it and stayed unconvinced, so he went out to disprove it. In April 1855, seven months after the last death, he traced the chain back to the baby at number 40, a case Snow had never counted, as S.P.W. Chave set out in Medical History.
The parish surveyor dug up the well and found the cesspool brickwork rotted through, inches from the well lining.
Whitehead also worked out what taking the handle off achieved. The baby’s father came down with cholera on 8 September, the morning the parish disabled the pump, and his waste would have gone the same way his daughter’s had.
Snow died in 1858, aged 45, with the profession still against him. In Florence that same autumn of 1854, an anatomist named Filippo Pacini had put cholera discharges under a microscope, seen comma-shaped organisms swarming in them and published. Italian medicine believed in bad air too, and Pacini’s finding went unnoticed by Snow, by Robert Koch and by almost everybody else until long after his death. Bacteriologists made it official in 1965, naming the organism Vibrio cholerae Pacini 1854.
Two men landed on the answer in the same year, in two cities, and the medical establishment believed neither of them. The handle came off anyway.
For $995 you could buy a decent second-hand lawnmower, a mid-range mattress, or, if a government spreadsheet has a bad day, a piece of Apollo 11 with moon dust still caught in its seams.
That last one happened. In February 2015, an online auction held on behalf of the US Marshals Service listed a white zippered pouch stamped “Lunar Sample Return”. Nancy Lee Carlson of Inverness, Illinois, bid $995 and won. She knew it had flown on something. What she did not know was that it had ridden along on the first crewed landing on the Moon.
How a moon bag ended up in a garage in Kansas
Blame starts with Max Ary, who ran the Kansas Cosmosphere and Space Center in Hutchinson from 1976 to 2002 and was convicted in November 2005 on twelve federal counts, including theft of government property, wire fraud and money laundering. Investigators searching his garage two years earlier had found a box, and the bag was inside it. As documented by The Hutchinson News, Ary was ordered to pay more than $132,000 in restitution for the space artefacts he had taken, a mix of property belonging to NASA and the museum itself. He always maintained he had simply muddled his own collection with theirs.
His forfeited property went to auction to cover the bill. That is where the mix-up happened. Per the Associated Press, government officials later said two lunar bags had been merged into a single entry when their inventory numbers were combined on spreadsheets. One was from Apollo 17. The other had been to the Sea of Tranquility, and nobody caught the difference before the gavel came down.
What the fabric was hiding
Carlson did the sensible thing and posted the pouch to NASA’s Johnson Space Center for authentication.
Terrible idea, as it turned out. Curators tested the cloth, found lunar material embedded in the fibres, and established that the pouch had served as the outer decontamination bag around the first rocks Neil Armstrong scooped up during the Apollo 11 moonwalk. Then NASA kept it, arguing the bag was government property that should be on public display, not sitting in someone’s living room, and asking the Justice Department to reverse the forfeiture.
Carlson sued. Twice, in two states.
The ruling nobody at NASA wanted
In December 2016, Judge J. Thomas Marten of the US District Court for Kansas ruled that Carlson was “a good faith purchaser, in a sale conducted according to law,” as reported by Robert Pearlman at collectSPACE. Marten went out of his way to say NASA had done nothing wrong here, and that the bag only mattered at all because of what NASA’s own people had pulled off in 1969. He also said he had no power to unwind a sale that had already gone through. Space.com later reported that NASA handed the bag back in February 2017.
Sold at Sotheby’s, well short of the estimate
Sotheby’s offered the pouch on 20 July 2017, 48 years to the day after Armstrong used it. NPR put the hammer price at $1.5 million, or $1,812,500 once Sotheby’s premium was added, which sounds enormous until you notice the house had estimated $2 million to $4 million. The buyer stayed anonymous.
Plenty of people thought the sale should never have happened at all. Michelle Hanlon of For All Moonkind, a group campaigning to protect the Apollo landing sites, said in a statement carried by Smithsonian magazine that the bag belonged in a museum where the world could share in the achievement it represents. Carlson told CBS News she decided to sell because she no longer felt it was safe in her home.
The dust went on to have its own career
NASA’s testing had lifted particles off the bag onto carbon tape mounted on aluminium stubs, and the agency held on to them after returning the pouch. collectSPACE reported that Carlson sued again in January 2019, claiming the testing had damaged the bag, dragged down its market value and caused her real distress, and demanding the return of the lunar material itself. That dispute settled out of court, and the samples went to her.
Which is how, in April 2022, five aluminium stubs, each topped with a 10-millimetre disc of carbon tape, came up in a New York saleroom. Bonhams sold them for $504,375, well under a pre-sale estimate of $800,000 to $1.2 million. Bidding opened at $220,000, seven bids came in, and the hammer fell at $400,000. Independent testing found four of the five stubs carried particles matching Armstrong’s first sample.
NASA still holds almost all of the 382 kilograms of rock, dust and core material brought home by the six Apollo landings. Its position on the rest hasn’t shifted in half a century: this material belongs to the public, not to private collectors. That stance holds up fine, right until two rows get merged on a spreadsheet.
Somewhere out there is a collector nobody can name, holding a cloth bag worth more than most houses, backed by a government receipt for $995 confirming the sale was done by the book.
A sailor burst into a small wooden railway station in the north end of Halifax on the morning of 6 December 1917 and told everyone inside to run.
Vince Coleman, the dispatcher on duty, walked back to his telegraph key instead. As the Maritime Museum of the Atlantic recounts, he was thinking about the overnight passenger train from Saint John, roughly 300 people aboard, due in at 8:55am on a line that ran straight past a burning munitions ship. He tapped out a warning to hold the train, then signed off: “Guess this will be my last message. Good-bye, boys.”
Minutes later the French cargo ship Mont-Blanc, packed with explosives bound for the war in Europe, detonated in Halifax Harbour.
A city with 2,000 dead and no spare hospital beds
Almost every building within half a mile of the water came down, and records held by the Massachusetts State Library put the immediate toll at 1,800 killed, 3,000 seriously wounded and another 10,000 needing medical attention. Later counts settle at close to 2,000 dead.
Halifax was in no position to absorb that. It was a wartime port, and as the medical journal Proto has documented, local hospitals were already carrying soldiers sent home to recover. Doctors worked through the first night and kept going.
Then it started to snow.
Boston did not wait to be invited
News crossed the wires and landed on the desk of Massachusetts governor Samuel McCall around 11am. What he had was fragmentary at best, a few lines about some sort of catastrophe up the coast. Writing in its account of that day, the Massachusetts Historical Society notes that McCall wired back a promise that his state would go the limit, and called an emergency meeting of the Committee of Public Safety.
By 10pm, twelve hours after the blast, a relief train loaded with doctors, nurses and supplies pulled out of Boston. Nobody in Halifax had yet told them what was needed.
A blizzard caught the train en route, hardly a surprise for Maine and New Brunswick in December. It did not turn back. Staff wired ahead for extra supplies while they were still stuck in the drifts, and the Halifax Municipal Archives documents the first American medical train arriving on 8 December, having pushed through snow for most of the trip.
The help kept arriving for years
That first train turned out to be the opening instalment. Author John U. Bacon, speaking to WBUR about his book on the disaster, counted two trains, two ships, 100 doctors, 300 nurses and about a million dollars in supplies, enough to stand up several temporary hospitals.
The medical unit went home on 14 December and formed the Massachusetts-Halifax Relief Committee on the journey back. The same archive shows it raised half a million dollars, refurnished more than 1,800 homes and set aside $25,000 for the care and education of people blinded by flying glass. Halifax named a block of temporary housing the Governor McCall Apartments, on a newly laid-out Massachusetts Avenue.
One tree in 1918, then a gap of fifty-odd years
Nova Scotia sent Boston a Christmas tree the following December as a thank you, and the tradition quietly lapsed.
It came back in 1971, and not purely out of sentiment. As reported by The Signal, the Lunenburg County Christmas Tree Producers Association revived the gift with one eye on the export market for Nova Scotian spruce. The provincial government took the job over a few years later and has handled it ever since, which is how a one-off act of gratitude became a line item.
What it takes to qualify as the tree
Anyone can nominate one. Criteria published by the Government of Nova Scotia ask for a white spruce, red spruce or balsam fir standing 12 to 15 metres, healthy, good colour, medium to heavy density, uniform and symmetrical, and easy to get a truck near. Officials narrow the field before checking their shortlist with Boston.
Boston’s 2025 tree was a 45-foot white spruce nominated by Ronald and Claire Feener of Lunenburg County. It reached the Common by flatbed truck under police escort, the City of Boston confirmed. Mayor Michelle Wu flew up for the cutting and, as CBC News reported, took a turn with the chainsaw herself on a 42-year-old spruce while a certified tree-felling instructor supervised.
That was a year of tariff threats and cross-border sniping, and the tree went south anyway.
Most international goodwill exists as paperwork nobody reads. This one gets scouted for months, felled in front of a crowd, driven hundreds of kilometres and switched on in a public park every December, all because a train once left Boston into a blizzard for a city that had barely had time to explain what was wrong.
Radium dust settled in their hair and on their coats, so a dial painter leaving the plant in Orange, New Jersey, trailed a faint shimmer down the footpath. Plenty of them thought this was marvellous. They painted their fingernails with it and dabbed it on their teeth before a dance. Nobody in management suggested they stop.
The paint was branded Undark, a mix of radium and zinc sulfide developed by Dr Sabin von Sochocky. In its account of the digitised company files, the National Archives describes how the Radium Luminous Material Corporation opened its Orange factory in 1917, selling the glow to watchmakers and instrument firms. Radium dials were prized in the trenches of the First World War and fashionable everywhere else.
How the paint got into their bones
Brushes were the whole problem.
Camel-hair bristles lose their point after a few strokes, and the numerals on a wristwatch face are minute. So the painters were taught to roll the brush between their lips before each dip. Lip, dip, paint, several hundred times a shift. The Science Museum Group notes that the New Jersey women worked on paper dials with a water-based paint, applied with fine camelhair brushes rather than the metal stylus used elsewhere in the industry. A trace of radium went down the throat with every dial.
Radium behaves chemically like calcium, so the body lodges it in bone and leaves it there, decaying quietly, for a life that just got much shorter.
The doctor who worked it out
Dentists noticed first. From about 1922 they were pulling aching teeth from young women in Orange, finding that pieces of jawbone came away too. A dentist, Theodor Blum, described the condition in 1924 and named it radium jaw.
Harrison Martland, the Newark pathologist who became Essex County’s first medical examiner, took it further. Writing in the Journal of the American Medical Association in 1925, he measured radioactivity in the tissue of dead dial painters and in the exhaled breath of living ones. A retrospective in Brain credits that work with pioneering the measurement of radioactivity inside contaminated human tissue.
U.S. Radium hired its own experts to challenge his findings. Reasonable enough in 1925, when one man’s diagnosis was still just one man’s diagnosis, though the company kept denying any link to radium for another decade.
Five women, one settlement, no admission
Grace Fryer needed two years to find a lawyer who would take her case. Katherine Schaub, Edna Hussman, Quinta McDonald and Albina Larice joined her, and by the first hearing in January 1928 none of the five could lift an arm to swear the oath.
They settled that June for $10,000 each, an annuity of $600 for life and their medical bills covered. The company admitted nothing, and most dial painters received no compensation at all, the National Archives notes. In 1936, a federal court in New Jersey ruled that U.S. Radium had not fraudulently hidden the danger, a decision that pushed surviving claims outside a two-year limitation period and effectively ended them.
What happened in Ottawa
Meanwhile, in Ottawa, Illinois, the Radium Dial Company assured its painters that the Orange trouble had nothing to do with them, since the radium over there was impure. The firm had been medically screening its own workers since 1925 and keeping the results to itself. As History.com documents, Illinois workers lost in court through the mid-1930s, after which the state passed the Illinois Occupational Diseases Act, effective 1936, requiring employers to carry cover for industrial poisoning.
Catherine Donohue led fifteen women in a fresh suit in 1937. By a hearing the following February she weighed around 32 kilograms and had to be carried into the room, and, as Northern Public Radio reported, she arrived with a box of what she said were pieces of her own jawbone. Local papers took to calling the group the society of the living dead, NPR Illinois reports, and their lawyer, Leonard Grossman, took the case for free because his clients had nothing to pay him. Later hearings moved to Donohue’s home, where she testified from a couch because she could no longer travel.
The Illinois Industrial Commission ruled for Donohue in April 1938 and confirmed the award that July. She died on 27 July, three weeks later. Radium Dial fought on regardless. The Illinois Supreme Court threw out the company’s petition to overturn the award in June 1939, and the U.S. Supreme Court declined a further appeal that October, more than a year after Donohue’s death.
That refusal to intervene left the award standing, and with it the principle that an employer could be held liable for a disease that surfaced years, sometimes a decade or more, after the job ended. Grossman’s son, three decades a Department of Labor lawyer, traces a line from those hearings to the federal OSHA Act, which he dates to 1971.
The measurements they left behind
Many surviving painters later volunteered their bodies to researchers at Argonne National Laboratory, whose review of the radium cases sets out how those measurements shaped radiation protection limits for the nuclear age.
Radium-226 has a half-life of roughly 1,600 years, so the women are still faintly radioactive in their graves and will be long after everyone reading this is gone. A surprising amount of modern workplace law rests on that fact. None of them were trying to found a legal doctrine. They wanted someone to pay the dentist.
A wind turbine blade is basically glass fibre set in hardened resin. The fibres give it strength, the resin holds them in place, and the two are bonded so tightly that a blade can survive wind, rain and stress.
That is the waste a group of researchers decided to work with. In a 2026 study published in ACS Applied Engineering Materials, a team at Xi’an Shiyou University, with collaborators at the Beijing Institute of Technology and Monash University, ground old blade material into powder and turned it into a lightweight, sponge-like insulation.
Here is what came out of the lab, and how it stacks up against the size of the problem.
It squashes to a fifth of its thickness and springs back
The most surprising thing about the material is how far you can crush it without wrecking it. The team reports it took a squeeze of over 80 percent and recovered, meaning you can press it down to roughly a fifth of its height and it bounces back. That matters because conventional silica aerogels are famously fragile. They insulate well but can be brittle. A material that can take a squeeze and spring back is far easier to handle and install.
The trick is in the structure. Instead of a solid block, the composite is a network full of tiny connected pores, with the blade powder stiffening a soft, springy scaffold.
It combined low density with unusually high compressibility
A caution is warranted: that line is the researchers describing their own work, not an outside verdict. “Outperforming” is their framing, and the comparison is against numbers from other published papers rather than a head-to-head test by a neutral lab. The figures themselves are measurable; whether the material performs as well in practice is the kind of thing other labs and larger-scale testing would need to settle.
It holds heat in about as well as ordinary insulation
Being light and squishy would not count for much if heat poured straight through it. It doesn’t. The composite measured a thermal conductivity of 0.068 watts per metre-kelvin, and lower is better here because it means less heat gets across. For comparison, ordinary glass wool building insulation sits around 0.04. So the blade composite is not a record-setting insulator, but it is in a sensible range for the job while remaining highly compressible.
It self-extinguishes instead of feeding a flame
Insulation lives inside walls and roofs, so how it behaves near fire matters. The team reports enhanced flame self-extinguishing capability, meaning it tended to stop burning once the flame source was removed rather than carrying the fire along. Again, this is one lab result on lab samples, though, and real fire-safety ratings require standardised testing. Read it as promising, not certified.
It turns a landfill problem into the raw material
The reason any of this is interesting is the sheer number of blades headed for the ground. Worldwide, the mass of blades expected to be discarded by 2050 may reach about 43 million tonnes. Right now, around 12,000 blades are retired across Europe and the US each year, and they are typically cut up and sent to landfill. A process that takes that same material and makes something useful flips the equation.
As the authors put it, reusing the blade waste “not only helps to reduce environmental burdens and resource waste but also opens up avenues for generating high-performing, useful materials.”
Where this actually sits
This is one study and one set of lab measurements. The chemistry is clever and the numbers look good, but the gap between a lab composite and a pallet of insulation shipped to a building site is wide. The paper doesn’t answer the questions that fill that gap: cost, batch-to-batch consistency, how the material holds up over years inside a wall, and whether grinding and processing old blades at scale actually makes economic sense.
That last question is the real hinge. Something has to make the recycled route cheaper or required before a lab result becomes a real supply chain.
None of that takes away from what the team built. This is a good answer to a narrow question, which is whether hard-to-recycle blade waste can be turned into a useful, light, tough insulating material at all. The evidence here says yes. The bigger question, whether it can be done by the tonne cheaply enough to matter, is still open.
What does it actually cost to throw away the future? For a retired solar panel in the United States, the answer is surprisingly small, and that small number is deciding where the panel ends up.
That gap is worth naming, because it isn’t really about whether recycling is possible. It plainly is. The gap is about who pays, and when, and that turns out to be a very different kind of problem.
Why the cheap option keeps winning
A solar panel is not a mysterious object. Roughly 75% of a crystalline silicon panel by weight is glass, with an aluminum frame, a junction box, silicon cells, and small amounts of copper and silver threaded through it. The trouble is that the parts worth money are the parts hardest to get at, and the part that’s easy to handle, the glass, is worth almost nothing on its own.
So the economics run backwards from what you’d hope. NREL’s own behavioral modeling of US modules through 2050 found that, under baseline conditions, about 80% end up landfilled, 1% reused, and 10% recycled. Landfilling wins not because anyone prefers it, but because it’s the option that clears at the lowest price on the day the panel comes down.
Contrast that with the European Union, where producer-responsibility mandates fold the cost of end-of-life handling into the product from the start. Under those rules, recycling a panel has been put at about $0.75 per panel, and EU recycling rates reach 95%. Same physics, same materials, wildly different outcome. The variable is policy, not chemistry.
What actually costs money when you take a panel apart
Pulling the aluminum frame and the junction box off a panel is straightforward. The expensive step is what comes after: cleanly separating the glass from the silicon cells, then extracting the tiny quantities of silver and copper that sit inside. Those metals are where the recoverable value lives, and they are present in vanishingly small amounts. Silver sits at well under a tenth of a percent of a panel’s weight, yet it drives the recycling math.
Because clean separation is hard, many operations skip it. They shred the whole panel and sell the low-value glass cullet, capturing a fraction of what’s there. And the prize for doing the harder work isn’t large: an electrical engineer writing in The Conversation estimates the recoverable silver and copper in a single panel are worth only about $10 to $12, which is why recycling loses money under current processes. Layers glued together make the glass reclamation harder still.
The field’s own practitioners have been candid about the limits. As of 2022, Arizona State University’s Meng Tao, who works on PV recycling, told Chemical & Engineering News that “The recycling technologies we have today are still rudimentary.” That was an assessment of a young field, and the field has moved since. The same Conversation piece notes that advanced methods like salt etching can now recover over 99% of silver and 98% of silicon at high purity. The capability is arriving. Whether the economics arrive with it is the open question.
Not everyone is sold on the crush-and-separate approach. Terry Ko, chief operating officer of PV Circonomy, whose company markets a layer-by-layer alternative, has argued that “Once glass, silicon, metals, and polymers are reduced into mixed particles, contamination becomes a significant challenge regardless of the subsequent separation technology employed.” Ko has a commercial stake in that critique, and the trade coverage it appeared in noted the absence of industry-wide purity standards. Worth hearing, not yet settled.
The retirement wave that’s already loading
All of this would be an academic argument if the panels weren’t coming. They are.
The technology to recover most of a panel exists, salt etching hits high-purity numbers, and a plant in Georgia is now claiming near-total value recovery. What keeps most panels in the ground is that landfilling is cheap on the day the decision gets made, and the cost of recycling lands on whoever is holding the panel at end of life rather than on whoever profited from it at the start.
The EU’s $0.75-per-panel figure is the tell. When the cost is built in early and the value chain is closed, the same object that looks like expensive waste in one system looks like feedstock in another. Cedartown’s bet on selling its recovered glass next door is a private-sector version of the same logic: make the low-value material worth something, and the math stops fighting you.
The $1-to-$5 number isn’t telling us that recycling is too hard. It’s telling us who currently pays for the easy way out. Change that, through mandates, through scale, or through a buyer waiting on the other side of the shredder, and the cheap option stops being the only rational one.
Bar soap moved a lighthouse. Crews smeared ordinary Ivory soap along the steel beams the tower rode across, because it turned out to be the ideal lubricant for a job involving thousands of tons of creeping brick, a detail recorded at the time by the Lighthouse Directory.
Everything else about the job was considerably less homely.
Why a lighthouse ends up in the surf
When Cape Hatteras Lighthouse was finished in 1870, it stood a comfortable 1,500 feet from the Atlantic. Then Hatteras Island did what barrier islands do, which is travel. Storm tides strip sand off the ocean side and dump it on the sound side, and the whole island shuffles westward, a process the National Park Service traces back at least 10,000 years.
The island moved. The lighthouse stayed exactly where the builders left it.
Underneath sat yellow pine timbers resting in fresh water on compacted sand, a floating foundation improvised because pilings could not be driven through hard sand barely eight feet down. Ingenious for 1870, and considerably less reassuring once the surf reached the door. By the late 1990s the ocean had closed to within 60 feet of the base, per the park service’s account of the light station, and decades of groins and sandbags had mostly made the erosion worse.
How to move five thousand tons of brick
Very slowly, and with an enormous amount of hydraulics.
Crews cut the tower free of its foundation, all 4,830 tons of it by the park service’s own retrospective, and replaced it with a grid of seven main beams, each 70 feet long and weighing 11 tons, fitted with 100 hydraulic jacks, as documented by Island Free Press. Those jacks raised the lighthouse 5.34 feet, just enough to slide roll beams underneath and give it a road. Horizontal push jacks then nudged the tower forward about five feet, backed off, shifted position along the beams, and pushed again, a sequence the park service sets out in its move FAQs. Each shove took about a minute, in coverage by the National Parks Conservation Association, and beams the tower had already passed over were carried round to the front to extend the track.
Progress ran to roughly 100 feet a day, with crowds cheering pushes that were barely visible to the naked eye, as Our State magazine later described the scene.
Twenty-three days, 2,900 feet, $11.8 million, and one American Society of Civil Engineers award the following year. International Chimney Corp of Buffalo held the contract. Expert House Movers of Maryland did the shoving.
The argument that came first
None of this was settled quietly. Relocation had been on the table since the National Academy of Sciences recommended it in the late 1980s, and North Carolina State University reached the same conclusion in a review published in January 1997, as Coastal Review later recounted, but plenty of locals were convinced the tower would shake itself to rubble halfway down the track. Congress funded the move anyway in 1998. The lawsuit that followed now sits in an administrative record held by the State Archives of North Carolina, a tidy paper monument to a very loud argument.
What the ocean has done since
The tower now stands about 1,500 feet from the water, the same margin it enjoyed in 1870, and should not be threatened again for something like a century, according to the State Library of North Carolina’s history project on the relocation.
Nearby property has fared worse. At least 32 privately owned beach homes have collapsed onto Cape Hatteras National Seashore since 2020, most of them at Rodanthe and Buxton, with the latest counted by National Parks Traveler in June. Each one brings a notice from Cape Hatteras National Seashore warning visitors about nail-studded debris drifting for miles down the beach. Superintendent David Hallac has pointed out that earlier generations built smaller and shifted buildings when the surf came for them, which is a polite way of saying the island never agreed to stay still.
Cracks the sea did not cause
Scaffolding went up in 2024 for what was meant to be an 18-month restoration costing $19.2 million, and the work has turned up something awkward. Crews expected 30 to 40 cracks in the cast-iron brackets carrying the gallery deck near the top. They found 116, spread across all 16 brackets, Stone & Lime superintendent Josh Brundige told a media tour in July, in reporting by Outer Banks Insider. Most were hidden under seven layers of brick, and the documentation of them predates the 1999 move.
Nobody can say what caused the cracking. Brundige offered thermal expansion and water damage as candidates, with generations of whale oil flame heating the lantern room and winter cooling it down again.
So the lantern room is currently missing from the tower altogether, a temporary beacon is doing the work, and the repair has added about a year and $10 million, taking the project to roughly $30 million. Moving a lighthouse five feet at a time was, in the end, the easy part. Whatever quietly split the ironwork at the gallery deck was already there, decades before anyone thought to look.
Concrete is the second most-used substance on Earth after water, and the cement that holds it together is one of the biggest headaches in climate policy.
That is the claim in a new paper out of India, and it comes with an unusual ingredient: human waste. A team turned treated sewage sludge into a fine black powder, swapped it in for a tenth of the cement, and got concrete that ended up harder to crush, not softer.
The problem with concrete is not the gravel or the sand or the water. It is the cement paste that binds them together. Estimates of cement’s climate footprint vary a bit depending on how you count, landing somewhere between 5 and 8 percent of global human-caused CO2 emissions each year. Most of that comes from the chemistry of making cement, not the fuel in the furnace. That is why switching to cleaner energy alone will not fix it.
One line of research has spent years asking whether you can replace some of that cement with biochar. Biochar is a carbon-rich char made by heating organic matter with very little oxygen. The appeal is twofold: you use less cement, and the carbon in the char, which the original plant or waste pulled out of the air, gets locked inside the building instead of released. The open question has always been whether the concrete survives the swap.
What the Warangal team actually did
The material here started at a sewage-sludge treatment plant in Warangal, in the Indian state of Telangana, a facility set up as a public-private partnership to tackle the city’s sewage problem. Human solid waste that would otherwise be dumped or burned gets processed there.
The study sets up a pointed contrast. Normally a city’s sewage sludge is a liability to be disposed of, and a builder’s cement is a material whose production loads the atmosphere. Both are problems flowing toward waste and emissions. The first is bigger than you might think: a single person produces up to around 400 grams of faeces a day, which across a city becomes a constant stream that has to go somewhere.
The idea here is to route that stream into the wall. The waste replaces some of the cement, and the carbon in the char stays put instead of being burned off. It is a tidy story on paper: two problems that might cancel out a slice of each other. The caveat is that this study did not actually measure the carbon side of that ledger, so the climate benefit is still an argument rather than a number the paper delivered.
A strong lab result is not a paving slab. Several things stand between this and a real building site, and the paper is upfront about the biggest one. Sewage sludge can carry heavy metals, and whether those slowly leak out of the finished concrete over time is an open question. That is not something you can wave away when the material might end up in a school foundation or a water channel.
How long it lasts is the other unknown. The tests ran to 91 days in controlled conditions. Real structures face decades of freezing and thawing, wetting and drying, weight and vibration, and nobody has shown yet how this concrete holds up over that span. The researchers say themselves that more work is needed to assess how it performs under real-world conditions such as freeze-thaw cycles and extreme temperatures.
What the study earns, we think, is attention rather than adoption. The explanation is plausible, the numbers are real for the mix that was tested, and the idea that a waste stream and a carbon-heavy material might each solve part of the other’s problem is worth chasing.
A tube of pond scum on a Cambridge windowsill spent half a year running a computer, and nobody had to feed it.
The container held blue-green algae, water and very little else, and it was roughly the size of an AA battery. Inside it, a colony of microbes harvested light, leaked electrons into a strip of aluminium and produced a current steady enough to keep a microprocessor ticking over from February to August 2021. Smithsonian magazine reported that the whole apparatus sat out a pandemic lockdown on a windowsill in the home of Cambridge biochemist Paolo Bombelli.
It ran far longer than anyone on the project expected.
What was actually inside the container
Synechocystis is a cyanobacterium, one of the ordinary non-toxic pond dwellers that has been quietly photosynthesising for a couple of billion years. Photosynthesis splits water and sends the resulting electrons down an internal transport chain inside the cell. Some of that current leaks out before it is used, and a metal anode positioned nearby can catch it. Exactly how those electrons make the jump from a living cell to bare metal is still not settled.
Bombelli and his colleagues, writing in Energy & Environmental Science, described a small housing of clear plastic and steel with an aluminium anode inside, built from inexpensive and largely recyclable parts. No feeding, no charging. Cyanobacteria make their own food as they photosynthesise, so light and water cover the entire supply chain.
Stranger still, the current did not stop at sunset. Researchers on the team suspect the microbes carry on metabolising food banked earlier in the day, which is why the cell kept nudging out electricity through the dark hours, as the Cambridge Independent reported.
How little electricity this really is
Averaged across the experiment, the cell delivered about 1.05 microwatts at 0.72 volts, figures Gizmodo set against the obvious comparisons. A fresh AA battery starts life at 1.5 volts. An efficient LED globe pulls around 10 watts. So the device was generating something in the region of a ten-millionth of a globe.
That sufficed only because the chip had been picked for extreme frugality. An Arm Cortex M0+ can tick along on 0.3 microwatts, and Arm Research built the test chip, the board and the cloud link that logged everything the microbes did. Only the chip itself ran on that current. The cloud link, which reported the results back to the researchers, needed conventional power to do its job.
Devices of this kind hold one advantage over the better-known microbial fuel cell, which needs a steady diet of organic matter to keep its bacteria working. Researchers publishing in Nature Communications noted that biophotovoltaic systems skip that input altogether, since the cells feed themselves as they go. The fuel bill is nothing and the colony renews itself, roughly the combination a sensor buried in a paddock for a decade would want.
Why a microwatt earns a paper
Who needs a computer powered by pond water? Nobody, right now. The argument is about the enormous, boring category of devices that comes next: door sensors, soil monitors, tracking tags, all the small gadgets multiplying across the world. Forecasts in the same paper put the number of connected devices on a path towards a trillion by 2035, every one of them wanting its own portable trickle of electricity. Batteries go flat, and they depend on materials that are expensive and unpleasant to dig out of the ground.
Christopher Howe, the Cambridge biochemist who co-led the work, argued in the university’s announcement that supplying all that hardware will require systems that generate energy “rather than simply store it like batteries”.
What one windowsill proves
Six months of unbroken output went into the submitted paper, and the hardware simply carried on afterwards. The funders at the National Biofilms Innovation Centre later described the microprocessor as having run for a year and counting on nothing but ambient light and water.
A single device in a single house is a demonstration, and one paper is one paper.
Two things stand between this and anything resembling a product: a great deal more current, and some agreement on how to measure it. A review in Frontiers in Microbiology characterises biophotovoltaics as a field hobbled by very low current outputs and by experimental set-ups too inconsistent to compare with each other, which stalls any systematic push to improve them. No one has fully pinned down how electrons cross from a living cell into an electrode either, so progress tends to come from trying things and seeing what sticks.
Scaling up means manufacturing thousands of these cells cheaply and reliably, and doing it before anyone has fully worked out what is happening inside them.
The processor was disconnected a long time ago. By Bombelli’s account the algae carried on making electricity regardless, entirely indifferent to whether anything was still listening.
In the IPU’s latest available ranking, Rwanda is first, ahead of Cuba at 57.2%. Around the world, women held 27.5% of national parliamentary seats as of 1 January 2026.
Rwanda isn’t narrowly ahead. It’s ahead by a distance.
The part that makes it interesting: the constitution doesn’t ask for anything like 63%. Rwanda’s constitution sets a floor of at least 30 per cent women in decision-making bodies. The country clears that floor by more than double, and keeps doing so, election after election.
How the 51 seats break down
The Chamber has 80 seats, and they aren’t all filled the same way. Twenty-four are set aside for women, chosen by electoral colleges drawn from local government structures. Another 53 are elected by universal suffrage under proportional representation, while two seats are elected by the National Youth Council and one by the body representing people with disabilities. The 24 reserved seats are exactly 30% of the chamber.
That isn’t where women’s representation stops. In the 2024 election, the IPU recorded 25 women directly elected and 26 indirectly elected. The indirect group included the 24 reserved women, one of the two youth representatives and the representative of people with disabilities. So women won 25 of the 53 directly elected seats — nearly half — on top of the guaranteed 24.
This isn’t a fluke from one strong year. Women held 48.8% of seats after the 2003 election, 56.3% after 2008, 63.8% after 2013, 61.3% after 2018 and 63.8% after 2024. The total has moved around, but it has stayed far above the constitutional floor.
The quota was not the only rule. A law on political organizations was later amended to require party lists to include at least 30% women candidates.
Why the total keeps landing above the floor
Part of the answer is simple math. The reserved seats guarantee 24 women no matter what happens in the general vote. Any woman who wins one of the 53 directly elected seats is added on top of that base. More than two decades of women serving visibly in parliament also means female candidates are no novelty, while the closed-list proportional system gives parties substantial influence over who appears in electable positions.
But the number is also tied to Rwanda’s wider political system. A quota can guarantee a minimum. It cannot, by itself, explain why women have repeatedly taken such a large share of the non-reserved seats.
The caveats the number hides
The headline figure describes the lower house, not all of power. As of 1 January 2026, women held 5 of 19 cabinet posts, or 26.3%, according to IPU and UN Women. Rwanda’s Senate was 46.2% women. A lower house that is 63.8% women therefore sits inside a political system whose other institutions are less female-dominated.
Then there’s the question of what kind of state produces a result like this. Rwanda is dominated by the Rwandan Patriotic Front, and Freedom House classifies it as “Not Free”. Its 2025 report noted that President Paul Kagame won a fourth term in 2024 with 99.2% of the vote after prominent opposition figures were disqualified.
Not everyone reads the numbers so darkly. A policy review on Rwanda’s quota points to reforms associated with women parliamentarians, including equal access to land ownership and equal-pay labour laws. Both things can be true at once: a genuine change in who governs, and an open question about how much independent political power those representatives wield.
So the number that opens all this, 63.75% against a 30% floor, is best read as a starting point, not a verdict. The reserved seats explain 24 women cleanly. They do not explain the 25 women directly elected in 2024, the long normalization of women in political office, or the way Rwanda’s tightly controlled party system shapes who reaches parliament. The distance between what the constitution requires and what Rwanda produces is exactly where the interesting questions live.
A war fought in orbit will not stay between the two countries that start it. That was the blunt message Gen. Chance Saltzman carried into his final public address as Chief of Space Operations, delivered to an international audience of military leaders at the Global Air & Space Chiefs’ Conference in London. Orbital conflict, the outgoing Space Force chief argued, is no longer hypothetical, and deterrence flows from combat-credible forces rather than declarations.
The speech capped a military career spanning more than three decades and a tenure leading a service that barely existed when he took command.
The core warning
Saltzman’s central claim was geographic as much as strategic. Satellites do not respect borders. A war fought in orbit — even one initiated between two adversaries — sweeps in every nation with assets overhead.
The message was aimed squarely at allied air and space chiefs. Saltzman framed space as a team sport — not by preference, but by physics.
Deterrence as capability, not signaling
Saltzman has spent his tenure pushing back against a version of deterrence built primarily on statements, red lines, and calculations about adversary intent. His alternative strategy asks the service to avoid operational surprise, deny adversaries the benefit of striking first, and develop counterspace capabilities that do not generate long-lived orbital debris.
The logic is that visible, credible force does the deterrent work on its own.
Saltzman argued that combat-credible capabilities should create a deterrent effect, and that if they don’t, the force will be prepared to respond to aggression.
That framing matters because it shifts the burden of prediction. Rather than trying to game out what Beijing or Moscow might do, the Space Force builds forces designed to defeat what those adversaries have already fielded — jammers, dazzlers, co-orbital interceptors, and cyber tools capable of disabling satellites.
The terrestrial precedent is already visible. Widespread GPS jamming across the Middle East during recent conflicts has disrupted commercial aviation, shipping, and civilian navigation across a wide region. Space-based services are already contested infrastructure. The question is whether the platforms themselves become targets.
Saltzman’s answer is that if adversaries believe an attack on U.S. satellites can be answered — and answered without turning low Earth orbit into a debris field — the attack becomes less likely.
A service transformed in four years
When Saltzman took command, the Space Force was less than three years old. It had a thin institutional identity and no coherent long-term force design.
By the time of his farewell, the service had grown and gained visibility through its role supporting recent operations.
The White House had requested substantial funding through the normal defense budget plus additional funding through reconciliation — a partisan-controlled maneuver rarely used for defense spending until recently. The reconciliation package was meant to fund programs including space capabilities and missile defense efforts. House leaders indicated they would back a significantly smaller reconciliation bill.
Saltzman declined to say publicly whether that shortfall would harm the service. He said he stood behind the request and called it a dramatic increase in resourcing for capabilities he considers vital. The rest, he acknowledged, is a negotiation between the executive and the legislature.
Alliances and the political weather
The London speech came after recent tensions between the United States and European allies. Saltzman’s message pushed the other direction.
He also offered a rare piece of advice to uniformed leaders operating in politically turbulent times, describing military institutions in democracies as providing stability — the weight that slows movement but keeps the vessel stable in a storm.
The framing is worth pausing on. Space is one of the few remaining domains where allied cooperation is not just useful but structurally required. Satellites cross national airspace hundreds of times a day. Space situational awareness depends on ground stations scattered across friendly territory. Debris from a single kinetic anti-satellite test can threaten spacecraft belonging to dozens of nations. The physics forces the politics.
What Saltzman leaves behind
Lt. Gen. Douglas Schiess is expected to become the third Chief of Space Operations if confirmed by the Senate. He inherits a service with a clearer doctrine than the one Saltzman received, but also with sharper political headwinds, a contested budget, and adversaries whose counterspace programs have matured considerably in recent years.
Acquisition speed, interoperability across allied systems, information sharing between classified and unclassified networks — these are persistent conditions of running a modern military service, not problems with endpoints. Treating them as solvable invites frustration. Treating them as manageable invites patience.
That’s a quieter argument than the headline warning about war in orbit, but it may be the more consequential one for the institution Saltzman built. The Space Force he leaves is younger than most of its guardians’ careers. It will spend the next decade discovering which of its problems are the kind that get fixed and which are the kind that get lived with.
For readers thinking about how space reshapes institutional identity and civilian imagination, our earlier piece on Artemis 2, Apollo 8, and overview thinking traces a parallel question from the exploration side of the ledger.
On December 14, 1972, in the last hour before Gene Cernan climbed the ladder of the lunar module Challenger for the final time, he crouched in the powdery grey soil of the Taurus-Littrow valley and drew three letters with his gloved finger: T, D, C. They were the initials of his daughter, Teresa Dawn Cernan. He knew, standing there in a place with no rain, no wind of the kind that moves dust on Earth, and no plate tectonics to churn the surface, that the marks he was making would very likely still be there long after the little girl who inspired them, and everyone she had ever known, was gone.
No human being has been back to touch them.
The last hour on the surface
Cernan and geologist Harrison “Jack” Schmitt had been on the Moon for more than three days. They had driven the lunar rover through a valley carved by an ancient impact, gathered more than 240 pounds of rock and soil, and, at one point, patched a broken rover fender using laminated geology maps and duct tape after Cernan snagged it with a rock hammer. It was the brightest moment of his life.
Then it was time to leave.
Schmitt was already inside the lunar module. Cernan lingered at the base of the ladder. He had spent part of the mission privately grasping for a line worthy of Neil Armstrong’s “giant leap.” What came out, into his suit radio, was quieter: “We leave as we came and, God willing, as we shall return, with peace and hope for all mankind.”
Before he turned to the ladder, he bent down. He traced TDC into the lunar dust with a fingertip protected only by the layered fabric of an EVA glove. He imagined, according to interviews he gave later, someone in a distant future finding the rover, the boot prints, and the three letters, and asking, “I wonder who was here? Some ancient civilization was here back in the 20th century, and look at the funny marks they made.”
Why the letters are still there
Earth erases almost everything. Wind sands down mountains. Rain dissolves limestone into caves. Roots crack pavement in a season. Ice widens fissures into canyons. A footprint on a beach is gone before the tide turns.
The Moon does none of that.
There is no atmosphere thick enough to move dust the way weather moves it on Earth. There is no liquid water at the surface. There are no plants, no burrowing animals, no plate tectonics reshuffling the crust. What does happen is slow. Micrometeorites — specks of interplanetary debris the size of sand grains and smaller — strike the surface constantly, gradually stirring the uppermost layer of regolith in a process planetary scientists call “gardening.” Solar wind sputters atoms off exposed grains. Temperature swings of more than 250°C between lunar day and lunar night flex and fatigue the rock.
None of that is fast. The top centimeter of lunar soil is turned over on a timescale measured in millions of years, not decades. The boot prints left by Cernan and Schmitt, and by the ten Americans who walked there before them, are, on any human clock, effectively permanent. NASA imagery from the Lunar Reconnaissance Orbiter has confirmed the Apollo tracks are still visible from orbit more than half a century later — a point Cernan himself made in congressional testimony, when he noted his footprints were still visible on the Moon more than 40 years after he had made them.
The same forces that preserve the Apollo hardware preserve TDC.
The claim that the initials will last millions of years is worth pinning down. The rate at which the lunar surface is churned by micrometeoroid impacts has been measured by studying the ages of exposed rocks and modeled from impact-flux data. Fine surface features roughly the depth of a fingertip trace — a few millimeters — are estimated to be erased on timescales of order 10 million years. Larger structures, like rover tracks and deeper boot impressions, last longer still. Meanwhile, cosmic dust rains onto every airless body in the solar system, including Earth, where it turns up preserved in unexpected places — in 2017, Imperial College London researchers recovered fossilised micrometeorites from roughly 90-million-year-old chalk of the same Late Cretaceous formation that makes up the White Cliffs of Dover.
On the Moon, that same slow dusting is one of the processes that will eventually bury the TDC letters. But “eventually” is not on a human scale. The letters will almost certainly outlast every printed book, every steel bridge, every marble monument on Earth. They will outlast the pyramids by a factor of a thousand. They will still be legible after languages that have not yet been invented have come and gone.
A father, an astronaut, and a nine-year-old
Cernan was in his late thirties on that last EVA. Teresa, whom the family called Tracy, was nine years old, in elementary school. His marriage to her mother, Barbara, would later end in divorce. He was, by every account, an aviator’s aviator — thousands of hours of Navy flight time, an electrical engineering degree from Purdue, graduate training in aeronautics, and a spacewalk on Gemini 9 that nearly killed him when his suit overheated and his visor fogged.
The image of a fighter pilot writing a child’s initials in the dust of another world is not the image Apollo usually offered the public. Apollo offered flags, salutes, and the pale-blue Earth hanging in a black sky. The astronaut office, as we’ve explored in a piece on Buzz Aldrin’s 1973 memoir, was a culture where emotional interiority was something you flew through, not something you talked about.
Cernan talked about it anyway. In interviews across four decades, he returned again and again to the moment on the ladder. “Those steps up that ladder, they were tough to make,” he told a NASA oral historian in 2007. “I didn’t want to go up. I wanted to stay a while.”
The TDC gesture sat inside that reluctance. It was a way of leaving something of his life behind in a place he suspected, correctly, he would never come back to.
Tracy grew up. The Moon kept the initials.
Teresa Dawn Cernan was a child when her father wrote her name on another world. She was in her fifties when he died in a Houston hospital in January 2017, aged 82. She is named in her father’s obituaries among his survivors, alongside her stepmother, Jan Nanna Cernan, two stepsisters, and nine grandchildren.
She has, on the rare occasions she has spoken publicly, acknowledged the initials as an extraordinary gift. It is also, quietly, one of the strangest inheritances anyone has ever received: not a house, not a watch, not a piece of jewelry, but three finger-drawn letters in extraterrestrial soil, protected from erasure by the physics of a world without weather.
The last footprint, and the reason it stayed the last
Cernan spent much of the rest of his life pushing, publicly and often, for humans to go back. He testified before Congress. He filmed a documentary. He worried aloud, more than once, that he would die still holding the title of the last man on the Moon.
He did.
The Apollo program ended not because the rockets stopped working but because the political reason for building them evaporated the moment the Soviet Union was decisively beaten to the surface. Apollo 18, 19, and 20 were cancelled. The Saturn V production line was shut down. Engineers who had built the largest rocket in history were laid off or reassigned. By the time Cernan retired from NASA in the mid-1970s, the infrastructure to send someone else back had already been dismantled.
More than half a century passed before anyone came close again. The Artemis program, now underway, plans its first crewed landing at the Moon’s south pole rather than Taurus-Littrow. Cernan’s valley is on the near side, at the eastern edge of Mare Serenitatis, more than 3,000 kilometers from the planned Artemis landing zone.
The initials will keep their solitude for a while yet.
A signature outlasting its author
There is a category of human gesture that only makes sense against the scale of geological time. Carving a lover’s name into a beech tree. Scratching initials into wet cement. Leaving graffiti in a Roman catacomb. What all of these share is a bet against forgetting — a bet the graffito will outlast the person who made it.
Most of those bets lose. Trees fall. Cement is repoured. Catacombs collapse.
The Moon is different. It is possibly the most stable writing surface in the accessible solar system. Cernan, an electrical engineer by training, would have understood this instinctively. He was drawing on a substrate that had been undisturbed for something like three billion years before he arrived. The mare basalts that underlie the regolith are older than most of the surface of Earth. When he wrote his daughter’s initials, he was signing a very old page.
In his final decades, Cernan gave a version of the same speech to schoolchildren over and over: humans belong out there. He was, by then, an old man in a blazer, talking to auditoriums of eight-year-olds. Somewhere in each of those talks, he mentioned Tracy’s initials. It was the human hook, the moment the room went quiet.
Gene Cernan died on January 16, 2017. His body is buried on Earth. The suit he wore in the Taurus-Littrow valley — the same one whose glove drew the letters — is a museum object now, gradually being studied and preserved by conservators who worry about the way lunar dust, once so alien it was quarantined on return, is slowly working its way out of the fabric.
Two hundred and fifty thousand miles away, on the eastern edge of a valley named for the Latin word for bull, three finger-drawn letters continue to exist in a place with no weather and no witnesses.
There is no fence around them. There is no plaque. There is no camera pointed at them. There is only lunar day, two weeks long, and lunar night, two weeks long, and the occasional silent flash of a micrometeorite hitting somewhere in the surrounding kilometers of grey.
If humans return to Taurus-Littrow in a century, or a thousand years, or ten thousand, and think to look, the letters will almost certainly still be there. If humans never return, the letters will still be there. If the species that made them goes extinct, the letters will still be there. If, on some very distant morning, another kind of visitor arrives from somewhere else entirely and finds a rover and a set of boot prints and, in the powdery grey soil beside them, three shallow marks in the shape of a T, a D, and a C, they will not know that they have found the way one father, on his last hour in another world, said goodbye to his daughter.
We are writers and editors reading the research, not clinicians, paediatricians, or child psychologists. What follows is reflection on published studies, not guidance for any individual family. The findings here are drawn from controlled research; population-level patterns are not prescriptions for your own child.
Feed some children sugar and others a sugar-free sweetener, and make sure nobody knows who got which. When you do that, the sugar does nothing measurable to how the children behave. That is the result pooled across roughly two dozen experiments in a 1995 review, and it has held up for three decades.
The children on sugar do not get more wound up. What changes, in a separate line of research, is the adult watching them.
Two dozen trials, no measurable effect
The anchor is a review published in JAMA in 1995 by Mark Wolraich, David Wilson and J. Wade White, then in the Department of Pediatrics at Vanderbilt University. Pulling together the controlled trials available at the time, they concluded that “sugar (mainly sucrose) does not affect the behavior or cognitive performance of children”. Reviewers note the analysis covered 23 experiments across 16 scientific papers, which is where the “two dozen” figure comes from.
This does not prove sugar can never matter. A small effect, or an effect in some particular group of children, could not be ruled out. What the pooled trials showed was no demonstrable change in behaviour, not certainty that sugar is harmless for everyone. The popular version tends to flatten into “sugar has no effect, full stop,” which the data does not quite support.
Why the trial design matters
The strength of the finding comes from how strict the rules were. To count, a study had to feed children a set amount of sugar, compare it against a look-alike sweetener, and keep everyone in the dark about which was which: the children, the parents, and the research staff all had to be kept from knowing.
That last part shapes everything. If a parent does not know sugar was served, they cannot pin behaviour on it. Picture a parent watching a child eat cake at a party and then act up: the parent has already decided what caused it. Take away that knowledge, and the effect disappears. Which raises the obvious question of where the effect was coming from in the first place.
The experiment where the parent was the variable
A study published a year earlier, in 1994, answered it about as directly as an experiment can. Daniel Hoover and Richard Milich, at the University of Kentucky, tested thirty-five boys, aged five to seven, whose mothers described them as sensitive to sugar.
The trick was in what the mothers were told. As the authors put it: “In the experimental group, mothers were told their children had received a large dose of sugar, whereas in the control condition mothers were told their sons received a placebo; all children actually received the placebo (aspartame).” Every boy got the sugar-free sweetener, but half the mothers were told their sons had been loaded with sugar.
The only thing that varied was the belief planted in the mother. And that belief moved the ratings. The authors report: “Mothers in the sugar expectancy condition rated their children as significantly more hyperactive.” How much trouble they saw tracked the story they had been given, not anything the child had eaten.
The behaviour that actually shifted was the mother’s own. The authors note that these mothers “exercised more control by maintaining physical closeness, as well as showing trends to criticize, look at, and talk to their sons more than did control mothers.”
This is one small study of thirty-five boys, so read it as a vivid demonstration of how the effect works, not a universal law about all parents. The direction, though, is hard to miss. Told a child had been given sugar, a mother hovered closer, watched harder, and found more to correct.
Why the belief outlasts the evidence
The 1995 authors had a plain explanation for why the belief persists despite the trials: expectation and simple association.
Sugar tends to turn up alongside the things that wind children up. Birthday parties, holidays, late nights, crowds of other children, a break from routine. The cake is present at the chaos, so the cake gets the blame.
An expectation, once held, tends to confirm itself in ways that are hard to spot from the inside. A parent who believes sugar causes trouble watches the child more closely after sugar, and closer watching finds more to flag. The child does not have to change for the report to change. Anyone who has scrutinised someone more carefully because they were warned to expect a problem has run a small version of this experiment without meaning to.
None of this means sugar is good for children. This is not an argument about diet or dental health, which are separate questions with their own evidence. The narrow claim is only this: under blinded conditions, the behaviour change so many adults are sure they have seen does not seem to show up.
The sugar-and-hyperactivity story may end up being less about children than about how confidently an adult can misread a moment once they already know what the answer is supposed to be.
If you are worried about a child’s attention or activity levels, we recommend you consult a paediatrician or child psychologist rather than making any decisions on articles such as this.
A kilometer under a mountain in central Japan, in a steel tank holding fifty thousand tons of water, physicists say they have caught the first faint sign of something that has never been seen before: the neutrino haze left by every massive star that has ever collapsed and exploded across the history of the universe.
The Super-Kamiokande Collaboration, an international group of roughly 250 researchers, reported the first indication of the diffuse supernova neutrino background. The result is real enough to be interesting and thin enough to demand caution. It is what physicists carefully call an indication, not a detection. The number that matters is 2.6 sigma, a confidence level of 99.5 percent, which sounds decisive until you learn that the field does not count anything as a discovery until it reaches 5 sigma.
So the accurate answer to the headline is: probably, for the first time, yes, with an asterisk that the scientists themselves are the first to insist on.
What the diffuse supernova neutrino background actually is
Every time a massive star runs out of fuel and its core collapses, nearly all of the energy of that explosion pours out as neutrinos, particles so slight and so reluctant to interact that they stream straight out through the dying star and off into space. Only a sliver of the energy emerges as the light we actually see. Across the whole observable universe, core-collapse supernovae go off several times per second. They have been doing so since the first stars formed.
The neutrinos from any single distant explosion are vanishingly weak by the time they reach us. But they never stop arriving, and they never fade. Over billions of years, the neutrinos from all those supernovae have piled up into a thin, permanent haze that fills all of space. That haze is the diffuse supernova neutrino background, and it is a genuinely different thing from the burst of neutrinos astronomers caught from a single nearby supernova in 1987. This is not one star. It is the summed exhaust of every core-collapse death in cosmic history, arriving all at once, all the time, from every direction.
Catching it would give astronomers a running tally of how often stars have been forming and dying since the early universe, a number they otherwise have to infer from starlight. That indirect count has blind spots. Some collapsing stars are buried in dust that hides their light, and some are thought to fail quietly, folding straight into a black hole with little or no visible flash at all. Neutrinos escape all of that, so the diffuse background should include deaths the optical surveys never register. It would also test what happens in the instant a stellar core becomes a neutron star or a black hole, an event that light can never show us directly because it happens behind a wall of collapsing matter.
Listening for a whisper this faint
The problem is that the signal is absurdly weak, and the tools for catching it were not built to be quiet. Super-Kamiokande is a 50,000-ton tank of ultrapure water watched by about 13,000 light sensors, buried under Mount Ikeno in Gifu Prefecture so that a kilometer of rock screens out the constant rain of cosmic radiation at the surface.
When a neutrino does interact with the water, it produces a brief cone of light called Cherenkov radiation, the optical equivalent of a sonic boom. The sensors read that flash and reconstruct what happened. For most of the years since Super-Kamiokande switched on in 1996, searches for this particular background came back empty in a specific way: they returned only upper limits, ceilings on how strong the background could be rather than any sign of it. The trouble was telling a genuine supernova-background event apart from ordinary radioactive and cosmic-ray noise, which swamps it.
The fix was chemical. In recent years the team dissolved gadolinium into the water, a metal that gives off a distinctive flash when it absorbs the neutron that one of these neutrino interactions leaves behind, letting analysts flag genuine events far more cleanly. The new result pulls together roughly 5,000 days of observation across two phases, one with pure water and a later one with gadolinium added. Sifting all of it, the team found an excess of events in the energy range from 13.3 to 81.3 million electron-volts, right where the diffuse background is expected to sit. The best fit points to a neutrino flux of about 3.6 particles per square centimeter per second. The team laid out the analysis at Neutrino 2026, the field’s main conference, in late June.
Short of the discovery threshold
Here is where the qualifier does its work, and it is not a formality. A 2.6 sigma result means that if there were truly no signal at all, random noise would produce an excess this large about once in every two hundred tries. That is enough to exclude the no-signal case at the 99.5 percent level. It is not enough to say the background has been found.
Particle physics learned to be strict about this the hard way. Fluctuations that looked at least this convincing have evaporated before, which is exactly why the discipline reserves the word discovery for 5 sigma, a threshold hundreds of times more demanding. Super-Kamiokande’s spokesperson, Hiroyuki Sekiya of the University of Tokyo, called the observation the world’s first indication of the background and a long-cherished goal of the project, and stopped there. He did not call it a detection, and neither should anyone reporting it.
Two more limits are worth keeping straight. The measured flux depends partly on models of how bright and how frequent supernovae were in the distant past, so the headline number carries theoretical assumptions inside it rather than standing entirely on its own. And the signal speaks only for core-collapse supernovae, the explosions of massive stars, not for every object the word supernova can describe. What the collaboration has, after all that, is the first statistical trace of a signal physicists have chased for three decades. Turning that trace into a measurement is the step still ahead, and the team is the first to say it has not been taken.
What a confirmed signal would open up
If the indication holds and sharpens into a detection, it would hand astrophysicists a tool they have wanted for decades: a direct census of stellar death reaching back toward the earliest generations of stars, read from particles rather than light. It would put hard numbers on how many collapsing cores end as neutron stars and how many vanish into black holes, and it would test the physics of the collapse itself against something other than theory.
The team is not waiting on chance. They plan to keep Super-Kamiokande running and to fold in its far larger successor, Hyper-Kamiokande, now being built to push sensitivity well past what the current detector can reach. Whether the whisper grows into a clear voice is a question for the next several years of data.
What the tank holds so far is a faint excess at 2.6 sigma: close enough to zero that a careful physicist will not yet call it real, and far enough from zero that the same physicist cannot look away.
The first X-ray ever taken of a human body, in 1895, was of Wilhelm Röntgen’s wife’s hand, wedding ring and all. One hundred and thirty years later, four amateurs orbiting Earth at roughly 17,000 miles an hour recreated that portrait — a hand, a ring, the same pale architecture of bone — except this image came down from space.
It is the first diagnostic X-ray ever captured of a human being in orbit, and it is now confirmed. A study published in Radiology, the journal of the Radiological Society of North America, reports that the crew of SpaceX’s Fram2 mission acquired anatomic radiographs during their 2025 flight, and that three independent radiologists on the ground judged the orbital images statistically indistinguishable from scans taken before launch — good enough to spot a fracture, a foreign object, or the kind of injury a crew doctor would need to see fast.
The device produced a full set of images across the multi-day flight: a phantom control object, a smartwatch, a hand, a forearm, a chest, an abdomen, and a pelvis. The scans were recorded digitally, so the crew could review them on the spot without developing film. The study was led by Dr. Sheyna Gifford, an aerospace medicine physician at Mayo Clinic, whose team had spent years working toward exactly this test.
Why X-rays, and why now
Space medicine has a physics problem. Ultrasound, the workhorse of orbital diagnostics for more than four decades, transmits sound waves through gel-coupled tissue. It works fine inside a pressurized cabin. It struggles with bone, where the internal structure often stays a mystery on ultrasound, and it does nothing for the interior of a cracked component. X-rays, being electromagnetic radiation, pass straight through.
That distinction matters as crewed missions push farther from Earth than at any time in half a century. In April 2026, the Artemis II crew traveled about 252,700 miles from Earth, breaking the distance record Apollo 13 set by accident in 1970. Lunar surface stays under Artemis III and the missions that follow will place crews days away from any hospital. A broken tibia on the Moon is not a broken tibia in Houston.
Four first-time flyers, four hours of training
The operational finding may matter as much as the clinical one. The Fram2 crew — cryptocurrency investor Chun Wang, filmmaker Jannicke Mikkelsen, engineer Rabea Rogge, and polar explorer Eric Philips — were all first-time flyers, and none were physicians. Three of them received roughly four hours of training on the portable system before launch, about the length of a morning workshop, then operated it themselves in microgravity.
That mirrors how portable X-ray already gets deployed on the ground. Gifford has pointed out that these units already run at the Kentucky Derby, on Super Bowl sidelines, and across low-resource regions worldwide, precisely because they can run on solar power and be operated by people with no medical training. Staying still enough to take a clean image in freefall was the open question, and the crew answered it.
The hardware also survived the ride. After launch vibration, re-entry heating, and a Pacific splashdown, the X-ray generator came back with only superficial structural damage. That is not a minor achievement. Most medical imaging equipment is designed on the assumption it will never leave a temperature-controlled room.
The Fram2 platform
The mission itself was an unusual choice for a first-of-its-kind medical test. The privately funded flight launched from Kennedy Space Center and became the first crewed spaceflight to fly a 90-degree polar orbit, passing from pole to pole roughly every 46 minutes at an altitude of about 265 miles. Over three and a half days, the crew ran 22 research experiments, from growing mushrooms to photographing aurora, with the X-ray work among the headline objectives.
The polar trajectory was incidental to the radiography. What mattered was the profile: a short-duration flight with untrained operators willing to serve as both subjects and technicians. That is close to the exact operational shape of a lunar sortie crew, and the International Space Station rarely offers it.
A dual-use tool
Before the orbital attempt, Gifford’s group had already flown a portable unit on a parabolic flight in 2022, showing that viable diagnostic images were possible in simulated microgravity. The orbital flight was the graduation exercise, and it extended the capability beyond bone.
The smartwatch image is the tell. An X-ray that can image a fracture can also image a suspect weld, a cracked pressure vessel, or a foreign object lodged where it should not be. Gifford has framed this as the point: for a sustained presence in space, X-rays matter not only for crew but for electronics and spacesuits, which cannot be inspected inside without either taking them apart or seeing through them. On a lunar habitat where every spare part had to be launched from Earth, non-destructive inspection is itself a life-support function.
The terrestrial implications may be larger still. Diagnostic imaging is one of the widest gaps in global health infrastructure. A device rugged enough for spaceflight, cheap enough to deploy at scale, and simple enough to run after four hours of instruction is a plausible answer to that gap — the same argument that already puts these machines in rural clinics far from any hospital.
What it does not solve
An X-ray is a diagnostic tool. It is not a treatment. A crew on the lunar surface with a confirmed compound fracture still faces the problem astronauts have always faced: the nearest orthopedic surgeon is a quarter-million miles away. The value of imaging in that scenario is triage — knowing whether to continue the mission, improvise a splint, or begin a medical evacuation that could take days.
Open questions remain. The images were taken in microgravity, not the one-sixth gravity of the lunar surface, where positioning a patient and holding source-detector geometry steady will behave differently again. Radiation exposure protocols inside a cabin already bathed in cosmic rays will need their own rulebook. And while the in-flight images matched the crew’s own preflight scans, the portable system’s ceiling against full hospital equipment is a separate question this flight did not settle.
The institutional signal
The broader story is about who is now doing this kind of work. A mission funded by a private citizen, flown on a commercial capsule, carrying a commercial off-the-shelf X-ray unit, tested by an academic medical center, produced a peer-reviewed result in a mainstream journal. NASA was not the operator. Mayo Clinic was not the launch provider. SpaceX was not the principal investigator. Each played one part, and the parts fit.
That distributed model is what lunar-era space medicine will probably look like. The Apollo-era approach — one agency, one contractor stack, one flight surgeon in mission control talking to one crew — cannot scale to the cadence of missions that Artemis, Gateway, and commercial low-Earth-orbit stations imply. A portable X-ray a non-physician can operate is a small piece of hardware. It is also a preview of the operational philosophy that will have to keep future crews alive when Houston is a light-second and a half away.
The hand that came down from Fram2 sits in the same lineage as the one Röntgen made in 1895. Same bones, same rings, the same faint gray shadow of tissue around them. One was taken in a laboratory in Würzburg. The other was taken by an amateur, hurtling over the poles, 265 miles above the planet, and it came home clear enough to read.
The Huygens probe drifted under a parachute through an orange haze on January 14, 2005, and set down on a plain of rounded pebbles that looked, in the first grainy image ever taken from the surface of an outer-solar-system world, exactly like a dry riverbed on Earth. The pebbles were water ice, cold-soaked to roughly minus 179 degrees Celsius and behaving mechanically like granite. The damp ground the probe splashed into was liquid methane. Titan, Saturn’s largest moon, was running weather.
Rain had fallen there. Rain would fall there again. And every drop of it was natural gas.
An Earth-shaped world with the wrong ingredients
Titan is the only body in the solar system besides Earth with standing liquid on its surface and a thick, weather-driven atmosphere overhead. The nitrogen in its air (about 95 percent, with roughly 5 percent methane) gives the sky a hazy tangerine cast and pushes surface pressure to about 1.5 times what a person feels at sea level in Miami. Radar had to be used to see through the haze at all, because visible light gets scattered into a permanent smog by photochemical byproducts drifting down from the upper atmosphere as fine dark soot.
Underneath that haze, the machinery of weather works the way a meteorology textbook says it should. Methane evaporates off the seas. It rises. It cools. It condenses into clouds. It rains back down. It carves channels into the ice bedrock, pools into lakes, feeds into seas, and eventually evaporates again. The methane cycle on Titan is a direct structural analogue of Earth’s water cycle, running on a molecule that on Earth is used to heat houses.
Bedrock cold enough to ring like rock
The ground itself is the part that keeps tripping up intuition. Titan’s crust is water ice, and at the surface temperature of roughly 94 kelvin (about minus 179 Celsius), that ice is not the slick, brittle stuff on a winter sidewalk. It is a stiff, load-bearing solid, harder than a lot of terrestrial stone. Mountains on Titan are ridges of frozen water. Boulders are chunks of frozen water. The pebbles Huygens landed among were frozen water, rounded by long tumbling in methane rivers the way quartz gets rounded in an Appalachian creek.
Recent modeling suggests the crust is not uniform, either. A team analyzing Cassini gravity and topography data concluded that Titan has an insulating methane-clathrate layer up to about ten kilometers thick sitting on top of the water-ice shell — a lid of frozen methane-and-water compound that traps heat from the interior and helps keep the atmosphere loaded with methane in the first place.
The seas that would float a boat
Titan’s northern hemisphere holds three seas large enough to have earned proper names: Kraken Mare, Ligeia Mare, and Punga Mare. Kraken alone is roughly the size of the Caspian, the largest inland body of water on Earth. Cassini radar sounded Ligeia Mare and found it more than 160 meters deep in places, deep enough that the radar echo from the bottom came back clean through a column of liquid so still and clear it startled the mission team.
The seas are mostly methane and ethane, with a little dissolved nitrogen. They are cold enough that a splash would freeze a bare hand into uselessness in seconds, but they behave, in bulk, like water. Waves have been detected on their surfaces, small ones, a few centimeters high, kicked up by Titan’s slow winds. Tides slosh through the straits between basins. And because the atmosphere is thick and the gravity is weak — about one-seventh of Earth’s — a person in a pressure suit could plausibly row a boat across Kraken Mare with a paddle.
Rivers that carved the ice
The most Earthlike images Cassini ever sent back were of Titan’s river networks. Dendritic drainage patterns branch across the terrain the way the Mississippi tributaries branch across a topographic map of Missouri. Some channels are hundreds of kilometers long. They meander. They braid. They cut canyons into the ice with steep walls and flat floors, and one system, Vid Flumina, was measured by Cassini’s radar altimeter as a canyon nearly 600 meters deep filled with liquid hydrocarbons — a methane fjord.
Not everything about Titan’s fluvial systems looks familiar, though. A recent SETI Institute analysis pointed out that Titan appears to be missing the river deltas that ought to form where its rivers dump sediment into its seas. On Earth, deltas fan out at nearly every major river mouth. On Titan, Cassini’s imaging turned up almost none. The reason is still being argued — sea levels may rise and fall too fast, tides may sweep sediment away, or the sediment itself may behave differently in cryogenic methane than sand does in water — but it is a reminder that the resemblance to Earth is a resemblance, not an identity.
Where the rain comes from
Methane rain on Titan is rare and heavy. Cassini watched thunderheads build over the equator and dump storms that darkened the ground for weeks afterward. The drops themselves would fall slowly — roughly 1.6 meters per second in Titan’s thick air and low gravity, about a sixth the speed of a raindrop on Earth — and they would be larger, close to a centimeter across. A person standing outside in a Titan storm would see fat gasoline-scented droplets drifting down through the orange sky at roughly the pace of falling snow.
The atmosphere is only about 5 percent methane by volume, but that is enough to drive the whole cycle. Sunlight, weak as it is at Saturn’s distance (about 1 percent of what reaches Earth), breaks apart methane molecules in the upper atmosphere. Those fragments recombine into ethane, heavier hydrocarbons, and the dark tholin haze that gives Titan its color. The haze drifts down and settles as sediment. That is why Titan’s beaches are black: the sky is slowly falling on them.
Why the atmosphere hasn’t blown away
By every quick calculation, Titan should not have an atmosphere. It is smaller than Mars. Mars lost most of its air billions of years ago. Titan holds onto a nitrogen envelope denser than Earth’s despite lower gravity and constant bombardment by Saturn’s magnetosphere. Its low temperature helps — cold molecules move slowly and are harder to strip away — and the methane keeps getting resupplied from somewhere below the crust, possibly from that methane-clathrate lid, possibly from cryovolcanic vents that have not yet been directly imaged.
Some of that history may have been violent. Modeling published through SETI proposes that Titan’s current orbit and spin were shaped by a catastrophic ancient impact that reset the moon’s rotation and helped drive the tidal migration that carries it slowly outward from Saturn to this day.
Lakes that might be building cells
The strangest recent finding is not geological but biochemical. A NASA study published in 2025 modeled the interfaces between Titan’s methane lakes and the tholin sediment that falls onto them, and found that cell-like vesicles can form spontaneously under Titan conditions. On Earth, the leap from chemistry to biology likely began when fatty acids folded themselves into little membrane-bound sacs — the first proto-cells. The NASA modeling suggests that the same trick works in liquid methane, using different molecules. Independent work has raised similar possibilities about bubble-like structures essential to life forming at Titan’s lake surfaces.
Nothing in that work says Titan is alive. It says Titan is one of the very few places off Earth where the raw geometry of a living cell might assemble itself out of local ingredients, in the local solvent, under the local sky.
What Dragonfly will fly into
NASA’s Dragonfly mission, confirmed and in development, will send a nuclear-powered rotorcraft to Titan’s Shangri-La dune field in the 2030s. The lander is a quadcopter roughly the size of a compact car, and it will fly because the air is dense and the gravity is light. It will hop from science site to science site, sampling dunes, ice bedrock, and possibly the floor of an ancient impact crater where liquid water and organic tholins may have mixed for tens of thousands of years — long enough, on the low end of prebiotic chemistry, for interesting molecules to appear.
Titan is one of the destinations where the vocabulary of Earth science works and the substance of Earth science does not. Rivers, yes. Rain, yes. Seas, yes. Bedrock, yes. Weather that runs the same cycle in the same order — evaporation, condensation, precipitation, runoff, back to evaporation — yes.
Just made of the wrong things. Every one of them. All the way down.
SpaceX is targeting Thursday for the 13th full-scale test flight of Starship, a mission that will carry Starlink satellites into a suborbital trajectory and attempt to resolve the two most consequential failures from May’s Flight 12: a Raptor engine relight in space and control instability on the Super Heavy booster during descent.
Flight 13 is the second outing for Starship V3, the redesigned vehicle SpaceX debuted in May with a taller propellant tank, upgraded avionics, and hardware for future orbital propellant transfer. It is also the first time the company will fly working Starlink satellites inside Starship’s payload bay rather than the mass simulators used on earlier tests.
What Flight 12 broke, and what Flight 13 has to fix
Flight 12 in May was billed as a mixed result. The upper stage reached its planned suborbital trajectory and executed a soft splashdown in the Indian Ocean. The Super Heavy booster did not. Booster 19 failed to stabilize itself for a controlled ocean landing after multiple engines encountered relight issues during the boostback burn.
The upper stage had its own problem. An engine anomaly forced SpaceX to skip the in-space Raptor relight demonstration, a capability that must work reliably before Starship can perform deorbit burns, orbital maneuvers, or the lunar descent required by NASA’s Artemis contract.
SpaceX attributed the booster’s issues to engine startup timing differences during stage separation. SpaceX reportedly stated that the vehicle demonstrated engine out capability and reached its planned suborbital trajectory, according to what was described as an official post-flight statement. SpaceX indicated that hardware and operational modifications were made to address the causes of the issues.
Flight 13 will attempt the same mission profile with those fixes in place. Ship 40 and Booster 20 will fly a suborbital arc lasting slightly over an hour, with both stages targeting ocean splashdowns rather than tower catches.
The heat shield is the real problem
Behind the engine relight and booster control objectives sits a harder engineering question that Elon Musk has repeatedly identified as the single biggest obstacle to making Starship economically viable.
Musk has stated that the single biggest remaining problem for Starship is having the heat shield be reusable, noting that no one has ever made a reusable orbital heat shield.
Starship’s belly is covered in thousands of hexagonal ceramic tiles that must survive repeated reentries without individual inspection or replacement. Falcon 9 has proven that first-stage boosters can fly dozens of times, but that vehicle never faces orbital reentry heating. Ship does, and it does so in a belly-first freefall profile more reminiscent of the Space Shuttle than a returning booster.
Musk has emphasized that if you want to be able to land the vehicle, refill propellant, and fly again, you can’t do laborious inspection of thousands of tiles.
To gather data on how the current tile configuration is holding up, several of the Starlink satellites on this flight will carry cameras aimed at Starship’s heat shield. The satellites will photograph and scan the vehicle’s tiles during ascent and cruise before deployment, giving engineers imagery that ground telemetry cannot capture.
Why Starlink V3 matters for the economics
The satellites aboard Flight 13 are the first operational V3 units to ride Starship. When Starship is fully qualified, each launch will be able to carry significantly more V3 satellites — a payload class that Falcon 9 cannot physically accommodate.
The capacity math explains why SpaceX has pushed Starship development so aggressively. A fully loaded Starship with V3 satellites would add substantially more throughput per launch than current Falcon 9 missions.
That expansion sits on top of an already dominant orbital footprint. SpaceX now operates more than 10,000 Starlink satellites, roughly two-thirds of all active spacecraft in orbit. Starship’s operational debut would accelerate that concentration rather than slow it.
NASA is watching, and running out of patience
Starship’s schedule pressure is no longer just commercial. NASA is counting on a Starship-derived Human Landing System to deliver astronauts to the lunar south pole under the Artemis program, and the agency has begun restructuring its plans around Starship’s slower-than-hoped progress.
Isaacman reportedly criticized the previous architecture as unsustainable or inadequate for achieving mission success.
NASA has asked SpaceX for a streamlined plan to speed the lunar return, and has also asked Blue Origin to accelerate its competing lander. China is targeting a crewed lunar landing by 2030, and both nations are aiming for overlapping locations at the south pole.
The stakes for the vehicle
Starship stands over 400 feet tall, making it the largest rocket ever flown. It is designed to lift more than 100 tons to low Earth orbit fully reusably — a capability no other vehicle, current or historical, has come close to demonstrating.
The vehicle still has a long list of milestones ahead of it even if Flight 13 succeeds: an orbital insertion, ship-to-ship propellant transfer, and the first catch of an upper stage back at Starbase. None of those are on the manifest for Thursday.
What Flight 13 can do is prove that the V3 architecture is stable enough to build on. A clean flight with a successful in-space Raptor relight, a controlled booster descent, and useful heat shield imagery would put SpaceX back on the trajectory the company sketched out before Flight 12’s engine problems.
A second consecutive failure of either objective would be harder to absorb. NASA’s revised Artemis timeline already assumes Starship recovers momentum quickly. Starlink’s V3 capacity buildout depends on the vehicle reaching operational cadence. And the underlying question — whether anyone can build a heat shield that survives reentry after reentry without a small army of technicians inspecting every tile — remains unanswered.
The launch window opens Thursday afternoon. Weather and range conditions permitting, the answer to at least part of that question will be visible from Starbase by nightfall.
In South Korea’s 2015 census, one in five people was named Kim: 21.5 percent of the population, or 10.69 million people. Next came Lee at 14.7 percent, Park at 8.4 percent and Choi at 4.7 percent. Add those four together and you approach half a country of roughly 51 million. Stretch the list to the ten most common surnames and you cover 63.9 percent of everyone, barely down from 64.1 percent in 2000.
Picture that in a Seoul subway carriage. A dozen people named Kim might be sitting together, none of them related. So why does a country of tens of millions run on such a short list of names?
The answer sits mostly in the last two centuries of Korean history.
Why the pool of names is so small
For most of the Joseon dynasty, a surname was a sign of rank, not something everyone had. Family names belonged to the yangban, the aristocracy. Commoners and enslaved people generally had none. The story of Korea’s few surnames is really the story of how everyone below the aristocracy eventually got one, and which names they reached for.
The spread began long before any law required it. A study of one prefecture’s census records, reported by Moon Ki-hoon in The Korea Herald, found that the share of local households without a surname fell from 45 percent in 1681 to just 6 percent in 1816.
The 1894 turning point and the rush to noble names
The legal end of the class system came with the Gabo Reform of 1894. When surnames finally became universal, the people newly free to choose one tended to pick from a small set of high-status clans. Kim, Lee and Park were the obvious picks, and for a clear reason.
Andrew Eungi Kim, a professor at Korea University’s Graduate School of International Studies, told ESPN that “those names were the most prestigious for having produced large numbers of kings, royalties and other historically significant figures.” In a society where family line shaped your opportunities, choosing such a name carried real weight. The top three surnames now cover 44.6 percent of the population, a share that has barely moved since the Joseon era.
Once taken, those names held firm. Yumi Moon, a historian of East Asia at Stanford University, explained the taboo against dropping an inherited name: “Changing your surname [which is] inherited from your ancestors is disgraceful, according to the custom of the olden.” A name chosen for advantage in one generation became, in the next, an ancestral inheritance nobody would think to change.
Living with a short list of names
So many shared names creates practical problems, and Koreans built a workaround into the system: the bon-gwan, or ancestral seat. Two people can both be named Kim yet belong to completely separate clans, each traced to a different founding region. The largest, the Gimhae Kim clan, has around 4 million members, with the Gyeongju Kim adding roughly 1.5 million more.
That distinction once mattered in law. An old rule banned marriage between two people who shared both a surname and an ancestral seat. In practice, millions of unrelated Gimhae Kim could not marry one another. The Constitutional Court ruled the ban unconstitutional on 16 July 1997. A revised Civil Code, narrowing the ban to close blood relatives, took effect on 31 March 2005.
Custom has proved harder to shift than law. Kim Beom-jun, a professor at Sungkyunkwan University, notes that “In Korea, the Confucian tradition still runs deep with people, an example of which is the practice of not allowing their children to use different surnames from those of their fathers.” He does not expect the mix to change: “I don’t think there will be many changes in the portions of family names in the future, either,” he said.
There is one quiet source of new variety. The 2015 census recorded 5,582 distinct surnames, more than 4,800 of them newly registered since 2000. Most came from naturalised citizens writing foreign names into Hangul. The list is growing at the edges even as its centre holds. Whether a handful of clans can keep anchoring the identity of a slowly diversifying country is the open question the census leaves behind.
On March 31, 2025, four first-time astronauts flying the first human mission ever to loop directly over Earth’s poles carried a commercial off-the-shelf X-ray machine into orbit — a wireless MinXray unit smaller than a briefcase — and used it to take the first diagnostic radiographs ever produced in space. The images were of each other’s hands, forearms, chests, abdomens and pelvises, plus a phantom calibration target and a smartwatch. They were rated by independent radiologists as equivalent in quality to the same crew members’ preflight scans on Earth.
The experiment sounds modest. Its implications are not.
Every serious plan to send humans back to the Moon, and eventually to Mars, runs into the same wall: what happens when someone breaks a bone, develops a kidney stone, or takes a hard fall on a dusty slope hours from the lander. Earth-based emergency medicine leans heavily on imaging. Off-planet medicine, until Fram2, did not.
What Fram2 actually did
The Fram2 mission flew commander Chun Wang, vehicle commander Jannicke Mikkelsen, pilot Rabea Rogge and mission specialist Eric Philips on a 3-day, 14-hour flight aboard the SpaceX Crew Dragon Resilience. It was the first human spaceflight to fly a 90-degree orbit, sending the crew directly over the north and south poles at an altitude of roughly 425 to 450 kilometers.
Tucked into the mission’s science manifest was a compact wireless radiography system — an FDA-cleared MinXray Impact Wireless generator paired with a flat-panel digital detector. Three of the four crew received four hours of preflight training on the device. None of the four are physicians.
Once in orbit, the crew imaged each other and a piece of hardware — a smartwatch, chosen as a stand-in for the kind of equipment inspection that could matter on a long mission — with no live guidance from the ground. Independent radiologists then compared the space-based images against preflight X-rays taken on Earth. Their verdict, published this month in the RSNA journal Radiology: the in-flight radiographs were equivalent to preflight images in overall image quality, spatial resolution and contrast resolution. Positioning for chest, abdomen and pelvis views was harder in microgravity and scored somewhat lower, but every image was rated diagnostic.
The generator came home on the Dragon with only superficial damage to its housing. Its internal hardware and X-ray output were unaffected.
Why bones matter more the farther you go
Astronauts on the International Space Station lose bone density in weight-bearing regions of the skeleton during extended stays. Fracture risk climbs accordingly, particularly during the physical shock of re-entry or the first days back under gravity. A lunar surface mission adds falls on uneven regolith, EVA-suit trauma, and equipment strikes to the list. Returning astronauts already describe pain in the soles of the feet and lower back within days of splashdown, before anyone falls off anything.
The ISS has managed without diagnostic radiography because it can be evacuated. A Soyuz or Dragon can bring a sick crew member home in hours. That safety net does not stretch to the Moon, and it certainly does not stretch to Mars.
The point was made sharply in January, when NASA cut short its Crew-11 mission on the ISS after a crew member developed an undisclosed but “apparently serious” medical condition. The agency canceled a scheduled spacewalk on January 7 and, the next day, announced Crew-11 would come home more than a month early — the first medical evacuation in the station’s 25-year history. The Dragon splashed down off San Diego on January 15. NASA officials said the decision was made in part because the diagnostic tools they wanted for the patient existed on Earth, not aboard the station.
On a lunar surface mission, that option collapses. On a Mars transit, it disappears entirely.
The training question
Deep-space crews will be small. Two, three, maybe four people. Not everyone can be a physician, and even if one crew member is medically trained, that person could be the patient. On Fram2, mission specialist Eric Philips is a professional polar guide, not a doctor.
The Fram2 result matters partly because it addresses that specific problem. The experiment demonstrated more than simply that radiography works in microgravity — it showed that non-experts, trained for a single afternoon, could produce diagnostic-grade images without live coaching from a flight surgeon.
The team behind the experiment, led by Dr. Sheyna Gifford, a Mayo Clinic aerospace-medicine physician, first tested the device on a parabolic aircraft flight in 2022 — the so-called Vomit Comet — which simulates weightlessness in bursts of roughly 20 seconds. Fram2 was the leap from a physics demonstration to an operational one.
Hardware inspection, and a dual-use dividend
The smartwatch image is easy to overlook. It shouldn’t be.
Long-duration missions fail more often from equipment problems than from medical ones. Cracked circuit boards, degraded seals, fatigued welds — these are the kinds of faults that ground engineers currently diagnose by talking a crew through visual inspection or waiting for a component to arrive back on Earth. A portable X-ray gives a lunar base the ability to look inside a failing part without disassembling it, or to confirm that a printed replacement matches spec. The Fram2 images of the smartwatch resolved internal components at submillimeter scale.
Space medicine has a long history of producing technology that ends up helping people who will never leave Earth. Portable radiography developed for austere spaceflight environments plugs directly into the needs of rural clinics, disaster zones and battlefield medicine — places where a fixed radiography suite is unimaginable. The MinXray system on Fram2 is already the sort of unit paramedics use at the Kentucky Derby and on the sidelines of the Super Bowl, as Gifford has pointed out. Space is just the newest austere environment.
What this does not prove
A few caveats deserve to sit in the open.
Fram2 was a short-duration mission in low Earth orbit. Radiation exposure was modest, thermal cycling was benign, and the crew was healthy. A three-year Mars mission looks nothing like that. Detectors degrade in high-radiation environments. Consumables run out. Software fails.
The imaging sample was also small — seven anatomic radiographs, on healthy crew members, taken with no urgency. Diagnosing a clean tibial fracture is one thing. Reading a subtle pulmonary embolism on a chest X-ray taken by a crewmate with four hours of training is another. The Fram2 result is a proof of feasibility, not a proof of clinical parity.
And X-rays are only one imaging modality. Kidney stones — a recurring risk in microgravity due to calcium loss — are usually diagnosed with ultrasound or CT, not radiography. A lunar clinic will need more than one instrument.
The direction of travel
Still, the trajectory is clear. Commercial short-duration missions like Fram2 are increasingly being used as testbeds for the medical and operational hardware that Artemis and its successors will depend on. Artemis II already carried its own layered biomedical research program on the crew’s April 2026 lunar flyby. The private spaceflight sector is becoming, in effect, an applied research arm for lunar and Martian mission planning — testing the tools that government programs will later fly to the lunar south pole and beyond.
A portable X-ray in orbit is a small step. It is also the sort of quiet, unglamorous engineering advance that separates a flags-and-footprints mission from a sustained human presence off Earth. Sustained presence requires the ability to treat injuries in place, inspect hardware without shipping it home, and give small crews real diagnostic tools instead of a satellite call to a flight surgeon and a hope for the best.
Chun Wang and his crew were in orbit for three and a half days. The little wireless generator they carried up with them, imaged each other with, and brought back down mostly intact, may end up mattering for decades.
Login
