Mechanical and robotic exoskeletons hold considerable promise, both as aids to the disabled and machines to increase the lifting power of worked in heavy industry, but so far the reality has lagged considerably behind the dream.

One of the principle obstacles faced by designers in the need for frequent recalibration of exoskeleton settings. Each system, of course, has to be tweaked to suit its individual user, but it must also be adjusted to accommodate changes in movement styles or speed as the user becomes tired or switches from one function to another. Although technically possible, such alterations, done in downtime by a technician, are costly and tedious.

Scientists at the College of Engineering at Carnegie Mellon University in the US, however, have developed an exoskeleton system that incorporates feedback mechanisms powered by the person using it, allowing it to self-adjust to changing mechanical demands in real time.

The researchers call the system "human-in-the-loop optimisation" and have published their findings in the journal Science.

Led by Dr Juanjuan Zhang, the scientists tested their new system by developing a ankle exoskeleton, suitable for use as either a prosthetic or to increase efficiency in jobs where lifting or climbing is a requirement. 

When adjusted for optimum efficiency the ankle apparatus reduced the wearer's metabolic energy consumption by around 25 per cent. The device was tried on a range of volunteers, all of whom were asked to move in 32 different patterns over the course of an hour.

"When we walk, we naturally optimise coordination patterns for energy efficiency," said team member Steven Collins. "Human-in-the-loop optimisation acts in a similar way to optimise the assistance provided by wearable devices. We are really excited about this approach, because we think it will dramatically improve energy economy, speed, and balance for millions of people, especially those with disabilities."