According to the 2011 census, the percentage of the disabled people currently residing in India is about 2 percent of the total population which, when converted to numbers, stands for over 26 million lives. With a decadal growth rate of 22.4%, disability is an ever growing problem and thus finding solutions for making the lives of these people better has become a task of paramount importance.

Archit Matta, Bhavya Kaushik and Ayush Kumar (From L-R in the picture) are pre-final year undergraduate students pursuing Industrial and Production Engineering at IIT Delhi. In their of quest of trying to solve a real life problem, they began working on a tool for hearing-impaired people, which would help them in identification of important alarming sounds such as car horns, safety announcements and emergency signals. The project has already been awarded the “Design and Innovation Translational Seed Grant” of INR 1,80,000 from MHRD, Government of India and has been carried out with the help of the Assistech laboratory, which is an interdisciplinary group of faculty, research staff and students engaged in using modern technology for finding affordable solutions for the disabled.

         Assistive Technology

THE MOTIVATION

The team set out to work with a broader aim of creating a device that could help the disabled sections of society. To identify a suitable and more specific problem statement, they sought the help of Professor P.V.M. Rao and upon his suggestion, decided to design an assistive device for those with hearing disabilities. Assistive technology refers to any item, piece of equipment, or product system, whether acquired commercially, modified, or customized, that is used to increase, maintain, or improve functional capabilities of individuals with disabilities.

In order to find out why the problem continues to exist despite multiple products being already available in the market, the undergraduate trio started interviewing interpreters and the disabled at the Association Of The Deaf, New Delhi. This was followed by collecting more information from officials of All India Federation Of The Deaf. They figured that the demand was of a cheap product which would be easy to use. One which could detect doorbells, the person’s name being called, the barking of dogs,honking of cars etc. With this clear picture in mind, and a thorough understanding of the existing assistive products, the team benched out a plan to create an inexpensive and more efficient device for the hearing impaired.

ABOUT THE PROJECT

There is no greatness in denying the good that assistive devices do to the disabled. But it was important for the team to develop a product that could actually be launched into the market and be used on a large scale. Given that more that 50% of the disabled population lives in the rural areas, the product could not be extremely sophisticated. Here are some salient features of the product:

  1. This device called the DEIASO, consists of one low cost band(LB) and an Android app integrated band (AB). The LB is a sound detector which is based on the loudness of the sound present. It has been taken into account that different types of alarming sounds have different amount of loudness and thus can be differentiated on this basis.
  2. The LB would detect sudden change in loudness in the user’s environment using a microphone, an audio amplifier and an Arduino.
  3. The Android app uses the microphone to monitor the sound and detect if there are specific sounds like car horns using machine learning algorithms. It then gives a vibratory and visual output on both the phone and the integrated band (AB).
  4. It classifies sounds based on MFCC – “mel feature cepstral coefficients” – which takes into account about 250 unique and varied features of sound clips such as amplitude, timbre and frequency.
  5. The Android app also come with Google speech to text feature.
                        The Arduino Circuit

 

                              Testing Data

TECHNICAL KNOW-HOW AND AID

The device majorly uses Arduino, Android development, PCB designing(Eagle) and Machine Learning. While most of the knowledge of these fields can be attributed to self-learning, the team also took the help of multiple Professors and the members of the Assistech Laboratory. Professor P.V.M Rao helped them narrow down the problem and form a basic roadmap to the solution. He referred them to Pulkit and Suman from the Assistech Laboratory who helped them with the designing of the circuit, the wiring and the designing of the product. Prof. Arun Kumar of CARE advised them regarding features and methods that could be used to differentiate various sounds. And finally, Shashank Gupta a student of M.Design at IDDC  helped in designing and creating the user interface of the android application.

                        Screenshot of the App

CHALLENGES FACED

Since the device aims to deal with emergency or alarming sounds, it was immensely important to create a product with an extremely quick response time. Thus, the team had to make multiple changes over the course of the project in order to lessen the response time. Moreover, since the device was meant to be inexpensive, there were additional constraint to the materials that they could use. The product uses an android app, but since rural areas would be more accustomed to cheap smartphones, the team also had to find a way to try and increase the relatively low computational power of these devices. Yet another problem was the localisation of the device such that it requires no internet connectivity. They had to design, fabricate and test the circuitry themselves. Soldering the SMD components on the PCB’s was also a tricky task.

              Various Iterations of the Design

WHAT’S NEXT?

Archit, Ayush and Bhavya have already proposed two versions with different features. One of them is to retail at INR 500 and has an instant response time of 5 ms even though it lacks the presence of an android app. The other version is slightly expensive and is set to retail at INR 1200. It comes with an android app and also identifies the sound in addition to detecting it. However, it’s response time is 0.4 seconds and more work is being done on its optimisation. The final user testing would be done in the month of December following which, the team will be able to release 30 market ready products for the disabled.

We would like to congratulate the team for their remarkable achievement and hope that their device can inspire the rest of IIT Delhi’s undergraduate students to tackle such real-life challenges. We are all being trained to become problem-solving engineers after all!  

 

LEAVE A REPLY

Please enter your comment!
Please enter your name here