IIIT-Hyderabad has the largest group in AI and machine learning in India: Prof. P.J. Narayanan

International Institute of Information Technology (IIIT) Hyderabad is a higher-education institute deemed-to-be-university, founded as a non-profit public-private partnership, located in Telangana, India. It is the first IIIT in India under this model.

Prof. P.J. Narayanan, Director of IIIT-Hyderabad, tells us more about the work being done at the institute. Excerpts from an interview:

DQ: Can you explain the unique PPP model that IIIT-H uses to self-fund its research?

Prof. P.J. Narayanan: IIIT was established in the Not-for-profit, Public Private Partnership model in 1998 for all its operations and for all infrastructure beyond the land and initial buildings. There are several educational institutions outside of the government system. However, most of them focus primarily on teaching and don’t have the wholesome, research-oriented institutional atmosphere like the top global and national institutions.

IIIT Hyderabad set out to be a research-oriented institution outside of the governmental system right from start. I am proud to say we are the pioneers in this and are far ahead of other institutions in the country. However, the model is not unknown in the USA. Top institutions like Caltech, CMU, MIT, Princeton, etc., have been following this model for several decades.

Funding research from own resources is very tricky. Research done at IIIT-H and other institutions are for the broad social good. Ideas produced in the universities reach the society through ideas, solutions, products, etc., over the course of time. Governments fund research in academia in all countries for this reason.

A significant portion of our research is funded through projects and grants from governmental funding channels of DST, MeitY, DBT, and others. Another big chunk comes from focussed government or public entities such as DRDO, DAE, PSUs, etc. The government is the source of such funds, but these are granted to the institute based on the competitive evaluation and assessment of the institute’s capabilities and track-record, with specific goals and deliverables. Funds are not guaranteed forever. We also get good funding from different industry sources today, again based on our track-record.

Lastly, we also use some part of the institute’s revenue through tuition fees, executive education, etc., to support research. The vibrant and cutting-edge research at the institute keeps its teaching current and our graduates highly sought-after.

DQ: How has IIIT-H's pioneering journey contributed to the growth of India's IT industry?

Prof. P.J. Narayanan: Academic institutions are like watering holes in jungles. Different players in the ecosystem benefit differently from them. IIIT-H has IT in its name. Our teaching programs are not limited to IT only, however, but may have one foot firmly in IT. Our impact on the IT industry is manifold.

The institute today produces about 500 well-trained individuals in core IT and allied areas annually. A big bulk of them join the industry in India while some go to industry abroad. Companies like Microsoft, Google, Amazon, Uber, etc., have a large number of our alumni as employees. A big company told us back in 2006 or 2007, that IIIT-H produced more high quality computer science graduates than all the IITs combined, and we hadn’t even completed a decade!

The research-oriented curriculum of IIIT-H also produces strong technology leaders; our alumni are present in disproportionate numbers in the core product groups of the large MNCs as well as the research groups of companies like TCS. The top-class talent from IIIT-H, I assume, played a role in the global MNCs expanding their India-based product operations. Facebook and Apple hired exclusively from IIIT-H in their very first year of campus hiring in India.

The institute also has been engaged in imparting current know-how to the industry in different ways. We provide opportunities for Hyderabad based IT professionals to take 1 or 2 courses at the institute without an elaborate admissions process. This Post-Graduate Student Status program started in 2005. About 3,000-4,000 professionals have taken advantage of this so far. We also have several offerings in the Executive Education space in current areas like AI/ML, Blockchains, IoT, etc., with online and on-campus components.

We are currently planning to expand our online presence significantly. We also conduct focussed, two-day programs for the top-leadership in companies to keep them updated. IIIT-H has done collaborative or sponsored research involving practically every company including the IT majors, Applied Materials, Goldman Sachs, Nissan, Honeywell, etc.

DQ: What was the pioneering journey, driven by the need for top-quality engineers to support India's IT industry?

Prof. P.J. Narayanan: Great educational institution have to be creators of ideas and technologies, beyond being trainers of technologists. IIIT-H focussed on training top-quality researchers first. Research is about exploration and unstructured problem solving. All our programs and all our courses have strong research-component in them. This trains the students in finding solutions by themselves. Several companies have appreciated this aspect of our graduates, starting with the initial batches.

We don’t believe in training someone for the first job they aspire for on graduation. We train students for the last job they take up 40 years after graduation. What skills and knowledge will be useful then? Learning-to-learn through research and exploration can last a lifetime. Curiosity and problem-solving attitude too can. Once students focus on these, the skills required for today become easy.

DQ: What are the insights that are shaping the tech landscape. What are the future plans for IIIT-H?

Prof. P.J. Narayanan: Being a research-focussed institution, IIIT-H, its faculty, and students are at the forefront action at all times. The institute has by far the largest group in AI and Machine Learning in the country, starting with the Language Technology, Computer Vision, and Data Science research groups over 20 years ago, when AI was not hot by any standard! That has paid of handsomely a decade later when AI became big.

We have also been tracking areas like Blockchain, Quantum Computing, etc., and have strengths in them. Multidisciplinary research with IT as one leg and areas like Sustainability, Natural Sciences, Human/Social Sciences, etc., are being pursued strongly at the institute. The institute is focussing strongly on Applied Research and Translational Research, whereby the fruits of academic research is taken directly to the society by working with the government, NGOs, start-ups, industry, etc.

DQ: What are some AI research-driven projects that focus on road safety?

Prof. P.J. Narayanan: The Kohli Centre on Intelligent Systems (KCIS) established by TCS in 2015 serves as an umbrella to coordinate and accelerate our activities in the broad AI areas. In addition, we have an Applied AI centre called INAI and a Technology Innovation Hub under the NM-ICPS by the DST. The major focus areas there are healthcare and transportation. In road safety, project iRASTE started in 2021 in partnership with the Nagpur Municipal Corp., CSIR-CRRI, Mahindra, as well as Intel and IIITH. The project fitted road-safety devices to about 300 buses that ply in Nagpur city.

The data from them are used to give alerts to drivers, to assess road conditions, and to suggest modifications to avoid accident hotspots. A similar effort was launched in Telangana on interstate buses that ply on highways in 2022. Both efforts have created large amounts of data and measurable impact on road safety.

DQ: Please mention a few AI-driven projects at IIIT-H, which include road safety, structural assessment of under construction buildings using drones, computational drug discovery.

Prof. P.J. Narayanan: IT and AI have wide applications. In addition to the above, IIIT-H has produced the Indian Driving Dataset (IDD) in collaboration with Intel, which has the largest road data in the world for unstructured driving conditions. This is a resource being used by start-ups and researchers in India as well as across the world. The robotics team has strengths in UAVs and drones and our computer vision team is strong in analysing visual data.

Combining these and with domain knowledge proved by our structural engineering experts, we have built solutions to inspect structures by flying drones with cameras. In healthcare, AI-based drug discovery is being worked on in our pioneering ML-based Science group. We are also working with Grace Cancer Foundation to screen large numbers of rural residents for oral cancer using just a mobile-phone camera. Societally useful applications of AI is enormous. We are eager to explore them in the coming years in collaboration with domain experts and users.