A woman jumped out of a seventh-floor window.

She was twenty-seven, 5-foot-3, and weighed 120 pounds. She broke through a pine-board roof on her way down, landing on her head at 40 miles per hour and lacerating her scalp. “The victim suffered abrasions over the dorsal portion of the spine and an oblique intra-articular fracture of the sixth cervical vertebra,” noted a 1942 report. The woman survived, and recovered in a hospital later that day.

The report’s author, Hugh De Haven, was intrigued that the roof had sustained more damage than the woman.

He went on to document seven other cases of attempted suicides or unplanned injuries in an effort to understand the physical limits and tolerance of the human body.

De Haven’s curiosity stemmed from an accident in 1916 when he was twenty-two years old. He had studied engineering at Cornell and Columbia Universities, after which he applied for a position in the U.S. Army Air Corps. Following a rejection, De Haven volunteered for the Canadian Royal Flying Corps as a cadet pilot.

One day during flying practice, De Haven was in a midair collision with another training aircraft. A 500-foot free fall ruptured his liver, gallbladder, and pancreas. His legs were fractured. De Haven wondered how he had survived when the same accident killed the other airman. Why did the same crash result in different injuries? This question laid the groundwork for the field of crash and survivability analysis that underpins the safety features of modern transportation systems.

Over the following years, to help make automobiles crash-proof, De Haven considered the principles of commercial packaging.

Boxes and containers are designed to withstand a variety of forces to protect their contents. As a basic principle, De Haven wrote, a “package should not open up and spill its contents and should not collapse under reasonable or expected conditions of force and thereby expose objects inside it to damage.”

From there, he built on the concept of “interior packaging” that would help prevent damage to the contents “from impact against the inside of the package itself.” And to achieve an optimal level of safety, De Haven added, a packaging engineer “would not test a packing case by dropping it [only] a few inches.”

Using a modular thought process informed by structure, constraints, and trade-offs, De Haven segmented automobile systems according to their safety elements: container, restraint, energy management, environment, and postcrash factors. The first letter of each of these elements creates the acronym CREEP, which offered a framework for studies on crashworthiness. De Haven’s literal comparison of passengers in an automobile to “fragile, valuable objects loose inside a container” eventually led him to patent the design for the three-point seat belt—now a standard automobile feature in most countries.

The seat belt design needed to differ from a shoulder harness – which De Haven knew from his experiences was effective for fighter pilots but not for automobile passengers. While the harness had the advantage of securing the upper torso and limiting “extreme forward movement,” it was uncomfortable and overly restricting.

De Haven’s belt could be comfortably adjusted across the lap and shoulder, minimising potential head injuries during a crash.

Seat belts help save tens of thousands of lives every year, substantially reducing deaths and injuries per mile travelled and thus enormously enhancing highway safety.

Let’s go back to De Haven’s studies for a minute. His subjects were voluntarily jumping out of windows and landing on their heads. So you might ask whether his studies were really “scientific,” and some people did call him a crackpot. Acceptable science relies on repeatable results. Purists may argue that De Haven’s subjects were anomalous and not representative of a “normal” population. Some of them were, in fact, trying to commit suicide but failed. I doubt if De Haven’s study protocols would ever be approved by an ethics committee under current laws. This wasn’t science in its pure form, but practice leading to evidence.

It’s difficult to “fully appreciate the fact that the head weighs as much as a ten-pound sledge hammer and packs the same terrific energy when it strikes a dangerous object at 40–50 mph,” De Haven wrote. “If the head hits a solid structure which will not dent or yield at such speeds, the head itself must yield, and crushing injuries of the skull and brain cannot be avoided. But if the head hits a light, ductile surface at such speeds, even a fairly strong metal surface will dent and bend and absorb the energy of the blow, thereby modifying the danger of skull fracture and concussion.” In 1946, De Haven went on to demonstrate in a famous experiment that a 11⁄2-inch-thick cushion would save eggs from breaking even when dropped from 150 feet.

De Haven’s observations show how trial-and-error engineering preceded organised science in giving rise to a new system of knowledge.

He reconfigured not only how we think about public safety, but also how we think about public health writ large. De Haven’s work effectively helped change the practice of seat manufacturers, airplane builders, and automobile designers in making seat belts an integral part of the safety systems in their products. Seat belts have been heralded by the U.S. Centers for Disease Control and Prevention as one of the ten greatest public health achievements.

Excerpted with permission from Applied Minds: How Engineers Think, Guru Madhavan, Penguin Books.