Stephen Hawking lecturing at KU Leuven University, Belgium, October 2011

Stephen Hawking may lose his iconic computerised voice

Professor Stephen Hawking could be at risk of losing his iconic computerised voice as the facial muscles he uses to control it are beginning to deteriorate.

The renowned cosmologist, who was diagnosed with motor neuron disease aged 21, is to celebrate his 70th birthday on Sunday (8 January). But the speech planned for the celebrations could prove to be one of the last he will ever give if he is unable to find a replacement voice production system.

Hawking currently communicates via an infrared sensor mounted on his glasses which relays the movements of his right cheek to a computer-controlled voice synthesiser.

But the degeneration of his facial muscles has progressed to such an extent that his speech has slowed to just one word per minute.

One possible solution is eye tracking, a technology currently being trialled by Letchworth-based company Time is Ltd (see E&T, October 2010).

Time is Ltd’s software set up as a face tracking system was featured on BBC’s Click programme last October. It uses facial recognition software to link the movements of a prominent feature, such as the nose, to a computer cursor. Eye tracking works on the same principle but as it focuses only on the movements of the eye it is a viable option for those with severely limited movement.

Speaking to E&T, Ian Schofield, Time is Ltd Director, said: “The system features a detector which follows the movements of the eye. This then allows the user to move a mouse cursor around the computer screen. We are currently working with a company in the States and will be getting the first example in the UK to try out with a MND patient by the end of the month.”

Schofield also believes the Cambridge University professor would benefit from switching to the company’s JayBee software.

He said: “With JayBee he would certainly be able to speed up his communication even without using the eye tracking system. It doesn’t use the full alphabet; it uses the kind of predictive alphabet found on a mobile phone, so instead of 26 options there are just eight.

“But even more important is the fact that JayBee intelligently predetermines what the user wants to say by learning not only the words but also the full sentences they frequently use.”

The A Brief History of Time author is notoriously reluctant to make changes to the system he has relied on since losing his speech in 1985 when a bout of pneumonia led to him having a tracheotomy.

However, Schofield is certain he would be able to adapt to the changes without any problems.

“I’m sure a man with the intellect of Stephen Hawking would be able to learn how to use it very quickly,” he said.

Further information

Pioneering speech-generating software for MND sufferers (E&T, vol 5 issue 17).

A voice for disabled people (E&T video, November 2010)

Recent articles

Info Message

Our sites use cookies to support some functionality, and to collect anonymous user data.

Learn more about IET cookies and how to control them