JOIN USWhat is the state of AI in India?
South Indian tourists seeking directions in broken Hindi from police constables belonging to Haryana, Uttar Pradesh and Punjab is a common sight around India Gate. Similar scenes are usually visible on the Mount Road in Chennai where tourists from eastern India seek help from traffic policemen on the road in English monosyllables.
This is one of the areas where Indian Artificial Intelligence research promises to chip in. Going forward, a day may come in the future when words spoken in one vernacular language would automatically be translated into another to ease out communication between people belonging to two distinct geographies.
“Machine translation across Indian languages and English would be an exciting area of AI research in India. The Ministry of IT is planning to start a Rs 400-500 crore project on speech-to-speech machine translation. It would involve advanced language processing and deep neutral network,” Pushpak Bhattacharaya, a professor at Indian Institute of Technology, Mumbai and principal investigator of the project, told SH. The partners in the project are IIT, Madras, IIT, Patna, IIT, Hyderabad and CDAC, Pune.
Several experts pointed out that India began late on AI and the availability of manpower is far too low. They also suggested it would be better if India identified a few select areas to focus on, so that a solution can be found to some of the vexed issues, rather than focus on many and fail on all fronts.
Keeping such an objective in mind, the Union Cabinet in December 2018 approved the Rs 3,660 crore national mission on cyber-physical system technologies that involves extensive use of technologies such as artificial intelligence, internet of things, machine learning, deep learning, big data analytics, robotics, quantum computing, quantum communication, quantum encryption (quantum key distribution), data science and predictive analytics.
The mission seeks to act as an engine to propel national initiatives in health, education, energy, environment, agriculture, security, infrastructure and industrial sectors. In the process, 40,000 jobs in the short run and 200,000 in the long run would be created.
“We will set up 25 hubs focusing on diverse areas. Each hub will have multiple academic institutions and industry. The call for setting up of the hubs is out and will be open for another month,” said Ashutosh Sharma, secretary, Department of Science and Technology.
The mission’s governing board is headed by former Infosys boss Kris Gopalkrishnan. Some of the other areas where AI could be of help are health, education, railways, aviation and metallurgy, said Bhattacharaya, who chairs several government panels on AI.
For instance in health, AI can help decipher an X-ray plate accurately, reducing the workload of overworked radiologists while in education, personalised curriculum can be framed for a child depending on his/her capability. The scope is evidently exciting and enormous.
However, limited availability of manpower is a major handicap. While China produces thousands of PhDs every year, the Indian output is not even in the hundreds.
There is also a shortfall in faculty. Availability of good quality, clean data is another challenge as there is no institutional mechanism to maintain high quality data. So even with government funding and industry participation, India is just at the starting point of a promising long road to harness AI for societal benefits.