About ColloquiumWe bring together leading researchers and academics, policy makers, and industry leaders

Artificial Intelligence has become ubiquitous today. Whether it is a recommendation for your next retail purchase or a conversation with a service chatbot or automated medical diagnosis, AI impacts us in our everyday lives. Every industry is striving to leverage AI and create competitive advantage in its business. At the same time, there are some key questions that are critical with respect to AI for India.

The IIT Madras Colloquium on AI aims to bring together leading researchers and academics, policy makers, and industry leaders to understand & deliberate on the state of the research and innovation in Artificial Intelligence (& Machine Learning) in India. It will raise four key questions:

1. What is the cutting edge research taking place in India? What should be our future research agenda?
2. How well is the research-to-market process working - success stories and challenges?
3. How do we get private sector & philanthropic donors to participate in scientific research?
4. What are the critical policy matters?

What's unique about this AI Colloquium?

The IIT Madras Colloquium on AI has a strong bias for action that is founded on our experience and data on AI from India. It aims to arrive at actionable programs for you / Indian AI ecosystem to contribute, collaborate and co-create.

- The Robert Bosch Centre for Data Science and Artificial Intelligence founded in August 2017, is one of the premier multi-disciplinary AI centers in India
- Our faculty chaired the Task Force on Artificial Intelligence for Economic Transformation set up by Ministry of Commerce and Industry
- Our knowledge partner, itihaasa Research and Digital conducted primary research to study the landscape of AI research in India. Download the report to learn about the six recommendations for strengthening the AI ecosystem in India

Share Your Ideas

Share interesting case studies / challenges / key policy issues in the Indian context corresponding to the Colloquium themes. Your ideas will be an input to the deliberations at the Colloquium



AI in EngineeringAnchor: Dr. Raghunathan Rengaswamy, IIT Madras

AI is increasingly transforming the world of engineering - whether it is to do fault diagnosis and preventive maintenance or combine with other technologies like IoT and cyber-physical systems to create smart factories of Industry 4.0. AI will transform the way any product or process is designed, manufactured and delivered, across all engineering disciplines.


The key questions that need to be deliberated in this scenario are

  • Are engineering facilities really data-rich ? Are we limited now by data silos due to the nature of the current solutions ? What is the state of digitalization that is required for ML/AI solutions to start making an impact ?
  • Are there engineering specific AI techniques that need to be developed - what are the current AI techniques that are of immediate relevance to engineering?
  • What are the developments that are required in enabling fields such as sensors, IOT, vision and so on that will make AI solutions effective?
  • What is the role of domain knowledge ? Is it the end-of-theory as some people are claiming ?
  • If domain knowledge has a role, how does one marry domain knowledge with data science/AI?
  • How important is interpretability in engineering vis a vis other domains ? How does one build this in ? What is the role of domain knowledge in this?
  • How does one build safe and reliable AI systems ? Are AI ideas for game-playing systems very different from the concepts that need to be developed for engineering?
  • How important is repeatability in AI solutions for engineering ?
  • How does one build truly learning systems for engineering applications ?
  • Engineering provides jobs to millions of people. What impact will AI have on these jobs and the society if humans are not required for these jobs anymore?
  1. Babu Narayanan
    Vice President AI, SymphonyAI
  2. Santoji Katare
    Technical Leader - Global Data Insight and Analytics, Ford Motor Company
  3. Ranganathan Srinivasan
    Fellow Engineer, Honeywell, India
  4. Venkoparao Vijendran Gopalan
    Sr. Gen. Manager and Head, Technology Strategy and Innovation, Robert Bosch Engineering and Business Solutions Ltd.
  5. Pankaj Doshi
    Head of Process Modeling, Pfizer Inc.
  6. Sanjay Vijayaraghavan
    Senior Engineer, GE Global Research
  7. Rajnish Mallick
    Principal Data Scientist, Rolls-Royce
  8. Karthik Ananthanarayanan
    VP and Head - Programs, Trucks, Ashok Leyland

AI and Ethics, Fairness and ExplainabilityAnchor: Dr. Harish Guruprasad, IIT Madras

AI is intended for the benefit of humanity. Right now, this technology happens to be a black-box. Unlike in traditional software systems where humans define the logic of computation, in AI systems we provide vast amounts of training data from which the system automatically learns and computes. Thus we do not know exactly how AI systems are arriving at their decisions. And their algorithms may be inadvertently influenced by human-biases and may thus adversely impact people's lives.

As AI systems mature, the expectations for global AI systems are becoming more complex, going beyond simple performance or accuracy on curated data. Hence it is very important to focus on ethics, fairness and 'explainability' of AI.

  • Ethics: Due to the multiple stakeholder nature of current AI systems, what is beneficial to the algorithm designer is often not so for others. e.g. recommending content to a person addicted to it.
  • Fairness: Systematic biases in data often make the AI system better for one group of people than others.
  • Explainability: The black box nature of current AI systems make understanding solutions difficult, and where answering "why" is as important as "what". E.g. explaining why the AI system diagnosed a patient with a particular disease is crucial, as it is finally the human who is the decision maker.


Some of the questions that need to be answered include:

  • How to be aware that a certain algorithm is a candidate for bias? How to define "fairness"? How to anticipate and weed out bias while building an algorithm?
  • What are the most prevalent methods right now for explaining a decision of an AI system to customers of such AI enabled services?
  • What are the "natural ways" for an AI system to communicate with a human?
  • How are companies in India preparing to uncover and address 'bias' in their AI solutions?
  • How can effective public policies and regulations be framed to address ethical considerations in AI solutions?
  • Are there learnings from how society has responded to ethical considerations in other technology domains?
  • What can the consumer do? Can AI/big data be used by the individual to find out and fix undesirable data policies that he/she is subject to?
  1. B. Ravindran, IIT Madras
    Professor and Head of RBC-DSAI, IIT Madras
  2. Vikram Vij
    Sr. VP and Head of AI CoE, Samsung R&D Bangalore
  3. Tushar Garimella
    Chief Growth Officer, Capital Float
  4. Bindu Ananth
    Chair, Dvara Trust & Head - Public Health, Niramai
  5. Muraleedhran VR
    Professor and Head of Centre for Technology and Policy, IIT Madras
  6. Manish Gupta
    CEO, Video Ken and Infosys Foundation Chair Professor, IIIT Bangalore
  7. Pulak Ghosh
    Professor, Decision Sciences and Centre for Public Policy, IIMB Chair of Excellence

AI for the next billion users in IndiaAnchors: Dr. Mitesh Khapra and Dr. Pratyush Kumar, IIT Madras

AI for India has some unique challenges, specific to the nation, its workforce, and economic/social conditions:

  • How do we develop AI solutions for India with uniquely high diversity in languages, literacy, infrastructure (low-end, high-end devices/network)?
  • How do we upskill a large workforce (IT, blue-collar, business) to move into an AI-first world?
  • How do we translate remarkable and fast-evolving AI research results to complex systems and domains such as agriculture, healthcare, judiciary, financial inclusion, weather prediction, sports, and e-governance?
  • How do we create a scalable and efficient platform for problem discovery, data sharing, and AI solutioning with appropriate incentives for academia, startups, industry, and the government?
  1. P Anandan
    CEO, Wadhwani AI
  2. Pushpak Bhattacharyya
    Director, IIT Patna
  3. Laxmi Prasad Putta
    Founder, Vassar Labs
  4. Sasisekar Krish
    CEO and Founder, nanoPix
  5. Gargi Das
    Director, IBM Research India and CTO, IBM India and South Asia
  6. Varun Gulshan
    Staff Research Scientist, Google
  7. Santosh M
    Commissioner, e-Governance at Tamil Nadu Govt and CEO Tamil Nadu e-Governance Agency
  8. Prateek Jain
    Researcher Microsoft Research

Program

Time

Session Name

Session Anchor

9.30AM - 10.00AM

Inauguration and context setting

Welcome address

Prof. Bhaskar Ramamurthi, Director, IIT Madras

Prof. Mahesh P., Dean IA&R, IIT Madras

Prof. B. Ravindran, Head of RBC-DSAI, IIT Madras

Kris Gopalakrishnan, Chairman itihaasa Research and
Digital and Chairman Axilor Ventures

10.00AM - 10.45AM

Colloquium Conversations + Q&A

Prof. Ravindran in conversation with
Kris Gopalakrishnan

10.45AM - 11.15AM

Tea / Coffee Break

 

11.15AM - 1.30PM

3 parallel Roundtable
Discussions

AI in Engineering

AI and Ethics, Fairness and Explainability

AI for the next billion users in India

1.30PM - 2.30PM

Lunch + Demos

 

2.30PM - 3.30PM

Overview and Take-aways from each of
the 3 tracks

Overview of the landscape study of
AI Research in India

Raghunathan Rengaswamy, Harish Guruprasad, Mitesh Khapra,

Pratyush Kumar Panda, IIT Madras

Krishnan Narayanan and N. Dayasindhu, itihaasa Research and Digital

3.30PM - 4.00PM

Colloquium Announcements

4.00PM - 5.00PM

Demos

Demos by Gyan Data, PadhAI, Resileo Labs, Buddi Health, GUVI

Registration

Thank you for your enthusiasm and ideas for the Colloquium. We are now closed for registration.
If you still want to attend the Colloquium, pl send a mail to info@itihaasa.com

Sponsors & Partners

Sponsor

Knowledge Partner

Knowledge Partner