Often, it is a challenge for visually impaired people to navigate through an unfamiliar indoor environment and find the desired location or public utility. Roshni, is the cell phone based system developed by a team of students from department of Computer Science and Engineering, IIT New Delhi along with the support of Saksham Trust, New Delhi and National Association for Blind, New Delhi which empowers the visually impaired people to navigate inside an unfamiliar building without any external sighted assistance.
Roshni is an affordable solution using infrared sensors which can be easily retrofitted to existing building infrastructure. The system is entirely controlled by the user and assistance is provided by the means of active audio messages coming from a mobile phone application, which also stores the map of the building.
When asked what prompted them to design such a solution Dhruv Jain, the team leader who completed his B.Tech from IIT Delhithis year, said, “According to the stats given by WHO, 87% of the blinds are in developing countries and India leads the race here. Presently there is no system available in India and other developing countries that allows a blind person to navigate freely in an indoor environment.”
“We incepted the idea and the design in May 2011, so it has been going for roughly 2.5 years now. The first prototype was completed in Nov 2011. We improved our designs over iterative feedback from users. The current system is the prototype version 3”, he added.
Talking about the extensive research for coming up with this effective solution, Abhinav Saksena, member of the team explained, “The development of such indoor navigation system posed many challenges for the team. We did an extensive research and followed a holistic development model applying insights from social and psychological dimensions into technology development. The iterative multi-centric user studies were necessary to draw conclusions and improve our user-centered designs.”
How it works
Roshni system has three components wall modules, a waist worn user module and a user interface which is based on a mobile phone application providing the directions and assisting the user to navigate inside a building. A network of wall mounted infrared sensors and accelerometers are installed on multiple floors of the building. Waist worn user module consists of IR sensors which interact with the wall modules and accelerometers to count the number of steps and an integrated bluetooth module to interact with the mobile application. The system can be easily installed in the buildings with minimum additions to existing infrastructure and the user can obtain directions to the concerned position, orientation and navigation just by pressing a key on the mobile unit, via audio messages.
Upon entering the building, the user turns on the waist worn unit and connects it to the mobile phone. Then the map of the building is downloaded in the mobile phone application. The map is in the form of an undirected graph with all the access points and meeting points of two corridors as nodes and paths between the nodes represent edges. The system follows Dijkstra’s algorithm and guides the user through the most convenient path to the destination. Convenience is preferred over path length, as user will prefer elevator instead of stairs even if the latter may provide a shorter path length.
User enters the destination in the mobile application via a QWERTY keypad. The application has an auto-complete feature to assist the user in entering the destination, it automatically predicts the text with the initial characters entered if there is only one destination present with those initial characters. User can any time change the destination by Cancel the Travel option and get updates about the current location using More Information on the Travel option. User gets all the information in form of audio messages generated using a text-to-speech engine.
As the user walks, user module constantly send updates to mobile application about the current position of the user and the number of steps data, dynamically calculating the path to the destination using the algorithm. Simultaneously, the user is also updated about the navigational directions to the next access point (viz. Intersection, T point, lift etc.) i.e., turn to take (left/right/straight) and the number of steps to travel. After reaching that access point, the user is again directed to the next access point and this process is continued till he/she reaches the destination. Wall mounted modules of IR sensors have integrated buzzers which beep when the user reaches his/her destination. Also, if the user takes a wrong turn, or deviates from the path, the application warns the user and the path is recalculated from the user’s current location to his/her destination. The device gives the instructions not only during wrong turns but also in the case of wide corridors and all this continuous confirmation of being on right path is conveyed through the vibrational alerts.
Roshii, is a user centric system which conveys the position to users as and when they want, making the building entirely transparent to the user, provided that map is installed in the mobile phone application and thus, giving complete independence to the visually impaired without any need of any external sighted assistance. Use of low cost IR sensor technology makes this system affordable for any moderate user in a developing country.
Any such system is unsuccessful untill the users believe it is useful for day to day living. Roshni, was tested with nearly 25 visually impaired persons over a period of 1.5 years and gained the trust of the users. They believed Roshni gives them confidence and expressed their willingness to use the device as a future product. Amit, one of the users said, “The device accurately conveys the navigational instructions, especially 1-2 meters before taking a turn. It warns me even when I do not follow the instructions and take a wrong turn. Initially, I got lost and took a long time to reach the destination. When introduced to the device, I was very optimistic. I gained confidence with the system during training and was able to easily reach the destination.”
Most of the users were comfortable with the user interface of the application and appreciated its features. Neha said, “Audible Indicator (buzzer on destinations) is a great feature. This helps me not to enter the wrong room or collide with obstacles. Also, pinpointing the location of doors is easier”. Raj, appreciated the audio message delivery of the application. Raj said,“Since the literacy rate is quite low in developing countries, the speech input can make the device universal to work with. The mobile application is user-friendly and the touch input is easy to learn.” Ritu added, “I feel aware of the surroundings just like someone is speaking to me.”
Roshni has got many recognitions and awards. Speaking about the awards and recognitions, Abhinav says, “Roshni has brought laurels to us. It has been awarded various awards in almost every field.
• It has been published at various places and has been recognized by the media.”
• Mobility and Transport for Elderly and Disabled People (TRANSED 2012).
• Best Poster Award, in the Poster Presentation at 4th IBM Collaborative Academia Research Exchange (I-CARE 2012) held in Indian Institute of Sciences, Bangalore, India
• Best project in the institute across all departments and disciplines, IIT Delhi, IIT Delhi Alumini Award with a cash prize of Rs. 2 lakh Winners in Eureka Paper Presentation, Techkriti, IIT Kanpur’s Technical Festival, 2013 [National Award] Project Incubation award in India Innovation Initiative 2013 [National Award] Wide media recognition in Deccan Herald, Hindustan Times, Times of India etc.
(EFY: For further information please visit: dhruvjain.info/me/Awards.html)
Speaking about the future, Dhruv said, “As of now, we are in the final development stages and our prototypes have improved a lot over these 2 years. Recent testing of the new prototypes with visually impaired users in IIT, Delhi has been very positive and encouraging.”
“We are currently improving the modules and also pilot testing our system at National Science Center, Pragati Madan, New Delhi. Our pilot tests are expected to finish by March 2014 following which we will start the full system installations. Our primary target would be the visually impaired users in India and secondary would be the sighted users. We will start our tests from public buildings and museums in Delhi and continue from there with constant improvements in the design”, he added.
Dhruv Jain, Research Associate, IIT Delhi & Research Intern, Microsoft Research, India
Abhinav Saksena, Research Intern, IIT Delhi
Akash Gautam, Research Intern, IIT Delhi
Nikita Gupta, Masters student, IIT delhi
Shiveesh Fotedar, Intern designer, IIT Delhi