How NASA Testing Changed US Food Inspection Forever

In 1906, in response to public outcry that came after Upton Sinclair’s The Jungle detailed the revolting conditions in Chicago’s meatpacking plants, the US government passed the Federal Meat Inspection Act. The legislation created what would become known as the Food Safety and Inspection Service, a force of federal employees who would spend their days in stockyards and packing plants, checking livestock for disease both before and after their slaughter. They did their work organoleptically—with their eyes, noses, and hands—an approach soon dubbed “poke and sniff.” This 100 percent testing method worked well enough for spotting the lesions that signaled tuberculosis and cysticercosis (a tissue infection caused by tapeworms), and the federal presence discouraged plants from using illegal preservatives and generally unsanitary practices. But by the end of the 20th century, as American taste for beef mushroomed, the system was clearly and woefully inadequate. The gotta-catch-’em-all approach was no match for plants butchering hundreds of animals an hour. The bigger problem was that testing had once more fallen behind in the arms race. Decades of poking at hearts, bile ducts, spleens, lungs, lymph nodes, livers, and kidneys had effectively rid the country’s meat supply of diseases that could be spotted or sniffed out. But humans can’t see, smell, or feel Escherichia coli. In 1993, the growing danger of E. coli made itself evident when hamburgers sold by Jack in the Box fast-food restaurants sickened more than 700 people and killed four children. The public outcry rivaled the one sparked by The Jungle and triggered two moves by the Food Safety and Inspection Service. As a short-term fix, it hired 160 extra meat inspectors. More importantly, it accelerated a change in approach that had been bubbling away in the background ever since NASA decided it was time to send astronauts to space for more than a few hours at a time—long enough for them to get hungry. That happened in the early 1960s, when the Mercury program, which first put Americans into space, gave way to Gemini, which proved out the equipment and procedures necessary for the longer, moon-bound Apollo missions to follow. Sending astronauts into space for multiple days meant dealing with the various elements of sustenance in zero gravity, from swallowing to defecating. But it was the middle step, digesting, that most worried Paul Lachance, the space agency’s lead food scientist. Food poisoning on Earth is terrible. In a spacesuit or space capsule, it could be catastrophic. The obvious approach would be a 100 percent testing route, screening every morsel for pathogens and taking a zero-tolerance attitude. This was unworkable, because, as with artillery shells, every bite of food that’s tested is one that can’t be eaten. This is known as “destructive testing.” Proving a given batch was absolutely safe would involve eliminating most of it. Private industry was no help; what inspection did take place in the 1960s was far from scientific or suitable for NASA’s needs. Most pathogens were “discovered” by the people who consumed them. Lachance and his colleagues responded by creating a new way of testing food safety. Where Walter Shewhart’s control chart was a tool for spotting manufacturing issues that could then be resolved, they conceived of an approach that went further, proactively identifying all the points where problems are most likely to start, and focusing inspection there. The core idea was to test food not after it was produced, but to so control its production that such testing was unnecessary. They called it Hazard Analysis and Critical Control Point, today better known as HACCP. To help digest this, think of basic acceptance sampling as the bit of wine the waiter has you taste so you know the bottle hasn’t spoiled before everyone gets a full glass. The basic Shewhart approach, then, is like the “test pancake” you make as you’re starting up breakfast. It tells you whether your production process, which requires a batter of the right consistency and a griddle at the right temperature, is on track. HACCP (pronounced hassip) is more like the way you should make a soup. Each time you add salt or broth or lemon juice or cream, you taste it to see how the flavor, saltiness, and acidity match what you’re aiming for. For NASA, analyzed hazards could be physical (crystallized Tang powder could break a tooth), chemical (the presence of pesticide residues), or microbiological (the pathogens that posed the most serious threat in space). A critical control point could be any moment or action in the production process, from gathering raw materials to vacuum-sealing a package. “We identified the critical steps that would make the difference between something being safe or not safe,” Lachance said. “The temperature, the pressure, whatever that criteria was, whatever that critical control point was.” A chicken headed into space would be plucked, deboned, cut into pieces small enough to freeze-dry, and cooked, at which point it would be sterile. But as various people handled it, the meat would be checked on its way to the freeze dryer, then on its way into plastic packaging, then again after being vacuum-sealed. To this day, no astronaut has gotten food poisoning while in space. “HACCP introduced careful design, rigorous standards, and precise measurement into the largely intuitive, loosely organized, hit-or-miss world of industrial food safety prior to the 1970s,” Timothy D. Lytton wrote in his 2019 book about foodborne illness, Outbreak. Pillsbury, which produced crumble-free food for NASA (floating crumbs were also a concern in space) and had a major hand in developing HACCP, implemented the process into its own consumer production lines after 1971, when a Connecticut woman found glass in some of its baby cereal. The approach went national in 1973, when the FDA added HACCP-based rules to its regulation of low-acid canned foods, where botulism was an issue. The European arm of the World Health Organization recommended using HACCP in 1983, and in 1985 the National Academy of Sciences declared it better than “random testing of foods” (aka inspection). In 1996, three years after Jack in the Box’s E. coli–speckled hamburgers made all those people sick, the Food Safety and Inspection Service demanded that America’s meat processing plants get with the program. By 2004, illness from E. coli had dropped 42 percent, reaching a Centers for Disease Control and Prevention goal six years ahead of schedule. In 2011, the Food Safety Modernization Act required that all US food manufacturers develop and stick to HACCP-based safety plans. Among experts, this way of thinking about food safety is as established as quality control is in manufacturing. “The testing of finished products for food safety is foolishness,” says Larry Keener, president and CEO of International Product Safety Consultants. He’s looking forward to the potential of new technology, including machine learning and genome sequencing, to make HACCP even more effective. But like all testing, HACCP’s usefulness depends on how it’s used. “Unfortunately, the industry and the regulatory community have, in my estimation, been slow to catch on and pick up and use this powerful strategy to improve the safety of the food supply,” Keener says. Indeed, every year, foodborne disease sickens 48 million people—that’s one in six Americans—puts 128,000 in the hospital, and kills 3,000. Those numbers are stubborn for a couple of reasons. Food safety is hard to see, so it’s hard for manufacturers to make it a differentiator. And while consumers may say they value safety above all, in reality price and convenience tend to win out, especially when it’s not clear what’s safe and what isn’t. The combined result is that manufacturers have little incentive to make more than the minimum effort. Meanwhile, the FDA (which oversees most food that isn’t meat) isn’t much of a task master. When it does run inspections, it’s focused on verifying that manufacturers can provide the proper documentation of their safety plan. “They’ll show up maybe once every couple of years. Or if a particular sector starts to cause a lot of foodborne illness, FDA will focus on that more. But usually it’s reactive,” says Denis Stearns, a founding partner at food-safety-focused law firm Marler Clark. “It’s only if you cause a big outbreak, then they also come in and do their own testing … They’re like the fire department. You don’t see them until your house is on fire.” Yet we have evidence that stricter standards and tighter testing can work. Since the 1960s, the French government has allowed certain manufacturers to slap the “Label Rouge” on their products, signifying that they have gone far beyond the legal minimum for quality and safety. It’s applied to all sorts of products, but is most closely associated with chicken. Its HACCP-like quality control process includes 65 individual tests, strict requirements on how much space and what kind of food the animals get, and—this being France—regular taste tests by experts and consumers to ensure the result is “vividly distinguishable” from the non-label version. Label Rouge chickens tend to cost twice as much as the regular birds, but make up one-third of sales and have a drastically lower salmonella rate than those in conventional flocks. Excerpted from Kobuk the Destroyer and Other Tales from the Wild, Unseen World of Test Engineering. Copyright © 2026 by Alex Davies. Used with permission of the publisher, W.W. Norton & Company. All rights reserved.