IN the weeks before he died, the universally-recognised voice of Stephen Hawking was given a new lease of life.
A project drawing on the electronics expertise of a University of Huddersfield student had succeeded in the difficult task of creating a reliable software emulation of the famous professor’s hardware-based text-to-speech system, which originated in the 1980s.
By early 2018, the software-based device was ready to be installed in Professor Hawking’s wheelchair. The sound quality was exceptional. “I love it,” was his reaction. Less than two months later – on 14 March – the professor died. His final words to friends and family had been spoken via the new voice emulator.
The long and complex project to upgrade the professor’s voice system was undertaken jointly by Peter Benie, a systems developer at the Engineering Department of Cambridge University, where the eminent professor was a director of research, and by Polish-born Paweł Woźniak, who is currently completing his Bachelor of Engineering degree at the University of Huddersfield.
Paweł’s involvement came when he landed a work placement with electronics giant Intel, which had sponsored and provided Stephen Hawking’s computer-based communication system since 1997. This gave Paweł an opportunity to pay visits to Professor Hawking and to repair two of the hardware boards that the motor neurone disease-afflicted scientist used to convert text to speech.
Also, the Polish undergraduate has made his contribution the basis for his 10,000-word final-year project at the University of Huddersfield, where his electronics engineering prowess has earned the praise of tutors. After graduation, he will take up a post in Cambridge with multi-national telecommunications equipment company Qualcomm.
Paweł and Peter co-authored an account of Stephen Hawking’s speech and communication system – available online – which includes its history, which can be traced to a device conceived in 1962 by an American professor as a way of enabling his blind daughter to convert text to sound.
The technology continued to evolve and in the early 1980s Professor Denis Klatt, of the Massachusetts Institute of Technology developed a system that included synthesised speech based on his own younger voice. This would be adopted by Stephen Hawking in 1985, after a tracheotomy had removed what speech he had left.
Over the years, Stephen Hawking’s support team included graduate assistants who had responsibility for maintaining and developing the Professor’s wheelchair and his speech and communication system.
It was one of these assistants, Jonathan Wood, who was in contact with Paweł Woźniak and the University of Huddersfield student was given the go-ahead to work on the voice system for his final-year project.
There were also a number of glitches that had to be sorted before the team gave Stephen Hawking a demo of the system, installed on a Linux laptop. To their relief, it was met with approval.
The software emulator behaves exactly the same as the hardware, said Paweł, and the Professor had been insistent that it retained exactly the same synthesised voice.
“But the difference was that it would not break down anymore, and the complexity of the whole system was significantly reduced,” he added.
“Another advantage is that the hardware was so old that it created a constant hiss. Whenever you spent time with the Professor you could hear it all the time in the background. In the software emulation, that was completely eliminated, so the voice was clearer.”
Paweł is grateful for the opportunity to work on a project with Stephen Hawking, and it was a great learning experience, he said. But the result was entirely a one-off development, he added.
Paweł commented: “I am proud to have had the chance to work on Stephen's communication system. I like to think that using his new voice gave him some enjoyment over the last few months of his life.”