Ability Forum

[srividyaa]

CHENNAI:
An ARM-based processor in an iPod can help play music; in a Nokia N97, it can connect to the cellular network; inside a Nintendo DS, the processor can make games come to life. Last Thursday, an ARM-based processor in a device called AVAZ helped Rohit Jain speak a full sentence for the first time in his life.

Rohit has cerebral palsy and cannot communicate orally. Like Santosh, who studies at Vidya Sagar [formerly Spastic Society of India, Chennai], and uses a chart to point at ‘H' and then at ‘I' to greet visitors. The two have also used devices like ‘Link,' which uses a speech synthesiser to voice out sentences they type on a keyboard, or ‘Pencentra' which converts their scribbled words into voice.

Rajul Padmanabhan, director, Vidya Sagar, says there are more sophisticated devices which cost around Rs. 3 lakh and which can help the kids voice their thoughts with speech synthesisers and accessible input systems, but there is no maintenance support available in India.

So when Vidya Sagar co-hosted ‘Silent Revolution,' a conference on Augmentative and Alternative Communication (AAC), together with the IIT-Madras, engineers were encouraged to look at developing an indigenous Voice Output Communication Aids (VOCA). Enter Invention Labs [www.inventionlabs.in], a company started by alumni of the IIT-Madras, and with support from professors of the institute, AVAZ began to take shape. “The availability of low-cost, low-power ARM-based processors and open source software was key to creating AVAZ at 10 per cent of the cost of similar devices available abroad. Even this cost is a function of the number of units produced and can come down as we go to mass production,” says Ajit Narayanan, a co-founder of the company, who came back to India after a stint in the United States in a technology company.

The user generates each sentence by scanning across letters and groups of letters on a touch-screen interface. A particular string is chosen by pausing the scanning process by touching the screen. Non-touch inputs are also recognised by connecting switches like ADITI (for motion sensing) or a mouse-shaped device (for tactile inputs).

At the end of the message, the constructed sentence is voiced out with a text-to-speech synthesiser. Ready-made templates are also stored for frequently used sentences, and the strings at each stage are also arranged based on the frequency of occurrence.

Kalpana Rao, a volunteer at Vidya Sagar, says AVAZ provides the children with an opportunity to communicate among themselves and with others. “With a chart, it is difficult for children in wheelchairs, say, to point out each letter to the other. When the words are voiced out, communication is much simpler.”

But she reckons that the device is not perfect. “We want full touch screen as those who can process information faster will be put off by the slow scanning process. They can directly point at the screen and form the words. We also want a local language interface.”

Ajit estimates that adding a full touch screen interface will increase the cost by around Rs.10,000. “But our software built using open source tools has been designed taking into account the frequency patterns derived from this model of use. We will need to change our software to reflect the full touch screen operations.”

He also stresses the importance of field testing the product. “After the IIT-Madras developed ADITI [a patented choice-activating device which uses motion data], they also gave us a number of ideas which we incorporated into their first piece of code. But only after coming to Vidya Sagar, did we find that we had to rewrite the entire software to suit the children's needs,” says Ajit.

As for local language support, he says the stumbling block is the lack of good text-to-speech synthesisers on which the company is working. Having obtained a Rs. 10-lakh grant from the Department of Science and Technology to develop the current ‘beta version' of AVAZ, the company is planning to apply for a further grant to get the other suggestions made by Vidya Sagar fulfilled. This will also help the company, as potentially all those lacking verbal communication can use the product, says company CEO Aswin Chandrasekaran. While those who can afford it will consider it a worthwhile investment, the company, along with Vidya Sagar, will look for potential donors for those who cannot afford it.

Source:http://www.hindu.com/2010/03/07/stories/2010030754401200.htm

CHENNAI:
An ARM-based processor in an iPod can help play music; in a Nokia N97, it can connect to the cellular network; inside a Nintendo DS, the processor can make games come to life. Last Thursday, an ARM-based processor in a device called AVAZ helped Rohit Jain speak a full sentence for the first time in his life.

Rohit has cerebral palsy and cannot communicate orally. Like Santosh, who studies at Vidya Sagar [formerly Spastic Society of India, Chennai], and uses a chart to point at ‘H' and then at ‘I' to greet visitors. The two have also used devices like ‘Link,' which uses a speech synthesiser to voice out sentences they type on a keyboard, or ‘Pencentra' which converts their scribbled words into voice.

Rajul Padmanabhan, director, Vidya Sagar, says there are more sophisticated devices which cost around Rs. 3 lakh and which can help the kids voice their thoughts with speech synthesisers and accessible input systems, but there is no maintenance support available in India.

So when Vidya Sagar co-hosted ‘Silent Revolution,' a conference on Augmentative and Alternative Communication (AAC), together with the IIT-Madras, engineers were encouraged to look at developing an indigenous Voice Output Communication Aids (VOCA). Enter Invention Labs [www.inventionlabs.in], a company started by alumni of the IIT-Madras, and with support from professors of the institute, AVAZ began to take shape. “The availability of low-cost, low-power ARM-based processors and open source software was key to creating AVAZ at 10 per cent of the cost of similar devices available abroad. Even this cost is a function of the number of units produced and can come down as we go to mass production,” says Ajit Narayanan, a co-founder of the company, who came back to India after a stint in the United States in a technology company.

The user generates each sentence by scanning across letters and groups of letters on a touch-screen interface. A particular string is chosen by pausing the scanning process by touching the screen. Non-touch inputs are also recognised by connecting switches like ADITI (for motion sensing) or a mouse-shaped device (for tactile inputs).

At the end of the message, the constructed sentence is voiced out with a text-to-speech synthesiser. Ready-made templates are also stored for frequently used sentences, and the strings at each stage are also arranged based on the frequency of occurrence.

Kalpana Rao, a volunteer at Vidya Sagar, says AVAZ provides the children with an opportunity to communicate among themselves and with others. “With a chart, it is difficult for children in wheelchairs, say, to point out each letter to the other. When the words are voiced out, communication is much simpler.”

But she reckons that the device is not perfect. “We want full touch screen as those who can process information faster will be put off by the slow scanning process. They can directly point at the screen and form the words. We also want a local language interface.”

Ajit estimates that adding a full touch screen interface will increase the cost by around Rs.10,000. “But our software built using open source tools has been designed taking into account the frequency patterns derived from this model of use. We will need to change our software to reflect the full touch screen operations.”

He also stresses the importance of field testing the product. “After the IIT-Madras developed ADITI [a patented choice-activating device which uses motion data], they also gave us a number of ideas which we incorporated into their first piece of code. But only after coming to Vidya Sagar, did we find that we had to rewrite the entire software to suit the children's needs,” says Ajit.

As for local language support, he says the stumbling block is the lack of good text-to-speech synthesisers on which the company is working. Having obtained a Rs. 10-lakh grant from the Department of Science and Technology to develop the current ‘beta version' of AVAZ, the company is planning to apply for a further grant to get the other suggestions made by Vidya Sagar fulfilled. This will also help the company, as potentially all those lacking verbal communication can use the product, says company CEO Aswin Chandrasekaran. While those who can afford it will consider it a worthwhile investment, the company, along with Vidya Sagar, will look for potential donors for those who cannot afford it.

Source:http://www.hindu.com/2010/03/07/stories/2010030754401200.htm

[Adverts]