Given that scientists can probe the secrets of matter and gaze into the depths of the cosmos, it is a wonder that we still have anything to learn about something as seemingly simple as the way we walk. Yet how and why humans evolved to walk on two legs — a feat that many take for granted — remains an active field of research. Not least because no one has yet managed to build a humanoid robot with the natural grace and poise of, say, someone walking to the post office, let alone with that of an Olympic athlete or catwalk model.

Armchair scenarios to explain how humans became bipedal turn up occasionally as submissions to this journal. Yet most fail to account for the manifold adaptations that humans require to assume the posture that most toddlers do without thinking.

Every part of our bodies is adapted to bipedality. Every nerve is tuned, every muscle responsive to the merest twitch of feedback required to keep something so intrinsically unstable vertical and in motion, and all without conscious thought.

This instability is key: the more unstable the object, the more manoeuvrable it is. Imagine the control systems that would be needed for a child to fly a modern fighter jet, and you begin to grasp the scale of the problem.

This explains why Nature is proud to publish a report this week that some readers might at first glance dismiss as quotidian engineering rather than pure science (S. H.Collinsetal.Nature;2015). The authors have created a mechanical boot fitted with a disarmingly simple spring-loaded device that augments the force of the wearer’s calf muscles, reducing the metabolic cost of walking by about 7%.

That might not sound like much, until one considers two remarkable things. First, the device is completely passive — it has no batteries and consumes no external power. Second, given that we assume that adaptations produced by evolution tend to the optimal, it is amazing that an improvement of this amount is even possible. Natural selection has honed human walking over millions of years, and yet there is still substantial room for improvement. Despite the way the phrase is sometimes used, natural selection is not a motive force for progressive improvement towards inevitable perfection, but an endless series of compromises as organisms adapt to new circumstances given the anatomies they have.

Humans come from a long line of creatures that mostly walked on all fours, supported by a horizontal backbone with vertebrae held under tension. Making the backbone vertical puts compressive loads on vertebrae, which can have painful consequences; lower back pain is the leading cause of worker absenteeism after the common cold, accounting for 15% of sick leaves and hundreds of millions of lost work days annually. Matters are even worse for pregnant women, and there is evidence that the spine has evolved further fixes to accommodate the enormous and sudden burden of carrying a fetus (K. K. Whitcome et al. Nature 450, 1075–1078; 2007). So perhaps it is no surprise that there are still energy savings to be made, given that humans are products of evolution and not of conscious design. If we were going to be truly efficient at getting around, we would have evolved wheels.

A passive exoskeleton for the lower leg is a far cry from the commuter jet packs imagined in 1950s science fiction. Still, if one is inclined to dismiss a 7% saving as modest, one might easily imagine circumstances in which such a margin could mean all the difference between immobility and freedom. A passive device such as this might be just enough, say, to help a person with impaired mobility to get out of the house. A person for whom a trip to the post office might at present be as out of reach as a jet pack.