Getting a firmer grip

Fingertip sensor technology is being used to develop a new type of prosthetic hand, claimed to be capable of accurately mimicking human movement and measuring grip force.

It is part of a three-year research project into intelligent sensor systems (ISS) being carried out at

by Prof Neil White, whose team is also working on intelligent keypads for possible security uses in automated teller machines (ATMs).

ISS is regularly used in industry and military applications, but

plans to take the technology into new areas.

The researchers aim to apply the technology to an existing prosthetic hand, called the Southampton-Remedi, in a bid to overcome some of the drawbacks with the present design.

'Some of the problems may seem trivial. for example, we have many sensors that use a lot of cabling to power and retrieve signals, which disrupts the mechanics of movement,' said White.

Light and flexible

The fingertip sensors will process information locally, minimising the amount of information transmitted to a controller and reducing the device's cabling and power usage. This will make the hand lighter and more flexible. It will also be capable of operating for longer than the six-hour battery time of most prosthetic controllers.

The sensors work by taking acoustic signal measurements, which enables the hand to control and slow slipping movements. 'The sensors will also be able to tell the difference between a plastic cup holding a hot beverage and a pint of cold beer,' said White.

The research is being undertaken with the help of a £417,000 EPSRC grant to the university's Electronic and Computer Science division, in a programme that runs from June this year until May 2009.

However, the medical industry won't have to wait until then for the first prototype. A prosthetic hand will be available for evaluation by this summer and will be offered to people who have lost limbs, to gauge their opinions on human interaction with the device, as well as its weight and functionality.

The team will be developing other projects simultaneously with the prosthetic hand research.

It is already working with a keypad hardware developer to test another of its projects, which aims to create intelligent keypads for ATMs, to help prevent fraud.

The keypad will take biometric readings not from a fingerprint scan, but from the way the user inputs his or her personal pin number.

By integrating pressure sensors within the keypad, the technology will calculate force and the rate of change of force, creating a bio- metric signature from a series of keystrokes.

Design challenge

The project has already presented the team with a number of interesting design challenges. It will have to consider how a person's behaviour — and thus their biometric signature — changes over time and in different weather conditions.

The system has advantages over other biometric measuring techniques such as iris or fingerprint scanning, because it will measure the information in a way that does not require the user to know that they are being analysed.

Another project expected to yield positive results for the team involves human-powered, wearable sensors capable of wireless data transmission.

The researchers envisage that body-sensor networks could operate as in-home patient monitoring systems. The sensors would be programmed to transmit patient parameters such as heart rate, temperature or blood pressure remotely, via a mobile phone or the internet, enabling a patient to rehabilitate at home.

White initially sees the sensors functioning within a medical environment, but he believes they would also have a sports science application in remotely measuring athletes' body behaviour.

The sensors would provide in-depth details about style, body rotation and acceleration, with the aim of improving performance.