Robotic technology is increasingly relevant to the entire healthcare sector. And the need for assistive robotic devices is set to increase dramatically as the demographic shift in world populations to older and more needy generations continues. Today, almost 20 per cent of the world's population is over 65 and this figure is predicted to exceed 35 per cent by 2050. Robotic devices such as prostheses, exoskeletons, rehabilitation and surgical systems will enable the elderly and disabled to regain their independence as well as maintain a quality of life they had never before thought possible.

Robotic feet and legs

Rock climber Hugh Herr suffered from severe frostbite while on Mount Washington and subsequently had both his legs amputated below the knee. Now he is professor at the Media Lab (MIT) and head of the Biomechatronics Research Group, a team that studies motion science and biomechanics to develop prosthetic knees and ankle-foots. Inevitably, Herr used himself as a study case and is now the owner of a robotic ankle-foot prosthesis.

Today's prostheses are built with advanced materials, but typically only provide a passive spring response during walking. This forces the amputee to have an unnatural gait and typically expend some 30 per cent more energy on walking than a non-amputee – think of it as a bit like walking with closed ski boots. Hugh Herr's robotic prosthesis has an electric motor which replaces the muscle, while a set of springs acts as the Achilles tendon. When the foot lands, the forward-motion energy of the wearer is stored in the spring and then released together with energy coming from the electric motor in a foot push-off action. An on board computer detects and adapts to the walking behaviour and the combined system reduces fatigue, improves balance and provides amputees with a more fluid gait. A single battery charge is sufficient to walk for a complete day, which equates to about 5,000 steps. The design of the prosthetic ankle also means that the whole contraption fits inside a shoe and under trousers, so the prosthesis does its job without any noise – and nobody even notices it's there. This incredibly successful replacement limb is now in the process of being commercially developed.

Herr also has another prosthesis, this time for his knee. The Rheo Knee from Ossur, consists of two magnetic plates separated by a magnetorheological fluid (MRF). This is made up of nanoscale iron particles suspended in a carrier fluid. A magnetic field across the MRF is created by an electric current crossing the plates, causing the iron particles to form chains. This increases the viscosity and in turn increases the resistance of the joint. A computer measures the load and angle to make the joint stiff during the stance phase to support weight and to allow for a smooth leg swing by making the joint compliant. It adapts to the walking style and environment, and is constantly learning to better itself.

Thought-controlled robotics

Even more intuitive control mechanisms are being developed, with prostheses being built that are controlled by thought processes alone. It's sufficient to only think about the job, as we would do naturally for our arms and hands, and so the robot is controlled. For this, the computer must tap the neural information from the brain or the nerves with several non-invasive and invasive techniques currently in existence. One non-invasive technique involves electrodes placed onto the skin above the muscle (electromyography – EMG) or on the scalp (electroencephalography – EEG), where in the latter the electrodes are usually fitted in a cap.

Jesse Sullivan lost both arms after accidentally touching a cable that contained 7,000V. A surgical operation rerouted the nerves that once controlled his arms through to muscles in his chest. Electrodes placed on the skin above these muscles now detect his desired motion and a computer converts the data to enact the actions of his bionic arm. So an electrical signal is sent to the hand to contract the muscle whenever Jesse wants to move the limb. The brain doesn't know the hand is gone, so the action seems perfectly natural when the signal contracts the muscle on his chest, which is subsequently picked up by the electrodes. The computer then receives the signal and thus opens the robotic hand.