India’s AI journey: The story so far

By Jibu Elias

Highlights

India ranks third in the world in terms of high-quality research publications in artificial intelligence. The same period also witnessed a staggering growth in AI startups in India with over 200 of them creating AI products and solutions.

Share this Article:

It was a hot summer day at Saidapet Government School in Chennai when Sethuraman, a Teach for India fellow, came across a hurdle that the education system has been facing for generations. The much flawed test-based student assessment system that survived decades fails to convey to the teacher what a student knows and what he or she doesn’t, realised Sethuraman from the disappointing result of a 6th-grade math test he just gave.

In search of an answer for students’ disappointing test performance, he held a thorough one-on-one with a student named Ravi (not the real name) who performed poorly on the test. To his surprise, he found out that the only thing Ravi didn’t know was how to subtract a three-digit number with borrow when zeros where there in the ones and tens places (eg: 500-198). Interestingly, understanding of this one fundamental mathematical operation was vital for answering many of the test’s questions.

The case of Ravi planted a crucial question in Sethuraman’s mind; how is he going to teach the students something new if he has no idea what they know, and they don’t know? And conducting a thorough one-on-one with every student in a class of forty was never going to be a practical option.

“If you want to go to Delhi airport and want directions from me, first I have to know where you are currently located at, as the directions from Noida to airport is different from Connaught Place to the airport,” Sethuraman explains. “Similarly, it is only by locating the learning gaps can one guide a student correctly,” he added.

Driven by this epiphany, Sethuraman along with Cibe Hariharan, his friend from college and an engineer with Amazon who has expertise in AI and machine learning, set about to invent a tool to solve this problem. And they found the “perfect alloy” to cast their new assessment tool in, a machine learning technology called reinforcement learning. In March 2018, they founded Jungroo Learning Solutions. 

Backed by a custom knowledge graph, Jungroo’s AI-powered student assessment system now uses many of the modern educational techniques such as personalised adaptive learning to help the teachers better understand the learning gaps in each student.

Today, Jungroo’s assessment tool is used by multiple educational institutions and organisations, including Chennai based NGO Bhumi. Bhumi deployed Jungroo’s software for the assessment of students in schools, orphanages, shelter homes, and slum areas - where they have been teaching.  

“Jungroo’s AI tool helped us understand the learning gaps in a student and streamline our process by solving volunteers’ data entry challenges,” says Satheesh Ravi, Program Manager at Bhumi. “It doesn’t stop with assessment, after that the application gives a personalised learning roadmap for each child.”

“It tells us what the kid knows, what the kid doesn’t know, and fear of assessment is gone amongst the students,” Satheesh added.

But this is not the story of Sethuraman, Bhumi, or the 6th-grader Ravi. It is the enabling technology called artificial intelligence or AI that empowered all three of them, as well as millions of others across the country to overcome multitudes of challenges across various sectors and industries.

Beginning of India’s AI journey

Artificial intelligence, as a concept and a field of research, was born out of a workshop at Dartmouth College in 1956 under the guidance of John McCarthy and Marvin Minsky. The idea of AI came to Indian shores first with works of Professor H.N. Mahabala in the 60s, and later with the creation of UNDP backed Knowledge-Based Computing Systems (KBCS) in 1986. But it was the US that pioneered and innovated in the field of AI. The last decade also witnessed an exponential advancement in AI technology as the data abundance of digital age led to the creation of machine learning. Soon China caught up with AI and set out targets to become the indisputable leader in global AI by 2030. All the while, India was seemingly dormant.

Yet, in recent years, there has been a steady rise in utilising AI in solving socio-economic challenges. This transformation is driven by entrepreneurs such as Sethuraman, the government, the broader IT industry, and academic institutes. Today, India ranks third in the world in terms of high-quality research publications in artificial intelligence. The same period also witnessed a staggering growth in AI startups in India. Currently, India has a booming startup ecosystem with over 200 of them creating AI products and solutions addressing challenges from fields such as healthcare, fin-tech, education to defence and national security. They have collectively received an investment of over 36 million USD.

Leveraging AI for India

When it comes to “AI transformation” India is at a nascent stage, and the government has been one of the torchbearers putting the building blocks in the right places. This is through funding various R&D projects, as well as starting schemes and initiatives across different ministries and departments to solve some of the tough challenges. For example, in January 2019, Centre for Artificial Intelligence and Robotics (CAIR), a DRDO laboratory, initiated a project to develop an AI-based solution for signal intelligence. The project will help to enhance intelligence collation and analysis capabilities for the armed forces. While on the side of governance, in January 2019, the Telangana State Tourism Development Corporation launched an AI-powered chatbot called T-Aadab. The bot provides on-demand data and information on tourism, landmarks, and tour packages. T-Aadab provides smart responses in all major global and Indic languages. 

Many of the central ministries and state departments are in the process of implementing various AI-powered project and schemes. A good example is how AI was used to help cotton farmers in Maharashtra. In the case of cotton farmers in Maharashtra, factors such as climate change, changing wind conditions, and soil composition have caused an increase in the variance of types of pest and the time of the month they attack the crop. This led farmers to resort to indiscriminately spray pesticide to prevent any loss due to infestation, often sabotaging their own harvests. As a result, the Maharashtra State Government entered in a partnership with the AI startup Wadhwani, which now has addressed this challenge successfully. 

“Wadhwani AI’s solution enables not just the identification of the pest but also helps the farmer or the extension worker make timely decisions. The solution aims to be integrated with existing government systems and serve about 80% of the 6 million cotton farmers in India,” according to Wadhawani’s spokesperson.

For India, AI is becoming the key for unlocking many of its socio-economic challenges. “Wherever there is a decision-making process, AI certainly has a role to play, and help in taking better decisions. If a proper decision is made, the life quality will improve, and that will help to address the socio-economic challenges. So in the Indian context, AI is a problem solver than a wealth generator”, says V Kamakoti, head of India’s AI task force and faculty at IIT- Madras. 

In the private sector, too, AI is now becoming a vital tool. A recent Accenture report points out that 80% of Indian corporate executives believe that AI is essential for not going out of business by 2025. Banking giant HDFC as well as retail-FMCG heavyweight Hindustan Unilever are all betting heavily on AI. HDFC’s EVA chatbot, built by Senseforth, interacts with thousands of customers every day with over 20,000 conversations from across the globe. Similarly, Hindustan Unilever’s AI software Jarvis runs predictive analytics on consumer behaviour. With the data collected from grocery stores, it can identify what customers are likely to buy next time.

India’s AI engine 

What is powering India’s aspiration to be one of the global leaders in AI is the country’s robust tech ecosystem, and booming startups, which are now using AI for addressing some of the long existed social problems in the country. Since 2011, the number of AI startups in India grew at an exponential pace of 86 per cent, beating the global average. In 2016, India was ranked third among G20 countries in terms of the number of AI startups after the US and China.

According to Kuldeep Patel, co-founder of Machinup, an AI a startup that provides predictive maintenance solution for industrial applications, “AI is the new Internet.” “The Internet transformed all the industries, including healthcare, education, and transportation. AI is going to do the same with time,” explains Kuldeep, as he explains India’s AI startup boom. “AI is no longer a luxury but necessity for industries in 2020, the sooner they adopt it, the better the impact will be,” he added. 

In the case of the IT industry, India’s IT titans are also building their own AI stack. Infosys’s Nia is an artificial intelligence and machine learning platform created to help businesses streamline data management and automate complex processes. Tech Mahindra’s GAiA platform and L&T’s Mosaic initiatives are other notable examples. 

Global tech giants such as Google, IBM and Microsoft have also been heavily invested resources in India’s growing AI ecosystem. Google last year launched its AI Accelerator in Bengaluru, while Microsoft launched AI Digital Labs in collaboration with ten higher education institutions to train 1.5 lakh students across India. 

Both of the “big tech” has also been partnering with multiple governments and private entities for using AI for social good as well. In 2016, Microsoft India, with Telangana Government initiated a cloud-based, advanced analytics pilot project to understand the health screening program of school students carried out by mobile health teams. The pilot project helped the government receive actionable insights by highlighting primary health conditions out of which vision impairment was found to be the most prevalent health issues among children.

In 2017, Google’s deep machine learning-based eye screening algorithm, that has achieved an accuracy of 98.6% in detecting diabetic retinopathy, was deployed in seventy-one vision centres across rural Tamil Nadu. This was in partnership with Aravind Eye Hospital to catch the early stages of the disease and prevent blindness.

Impact of ‘AI-future’ on India

The future impact of AI on Indian and global economy will be far-reaching as it is expected to double the annual economic growth rate of 12 developed countries by 2035. It could potentially deliver additional global economic activity of around $13 trillion globally by 2030, or about 16 per cent higher cumulative GDP compared with today. PwC estimates that out of this, $7 trillion is likely to go to China, $3.7 trillion to North America, and $957 billion to India. Which essentially will raise India’s annual growth rate by 1.3 percentage points, or 15 per cent of current Indian GDP.

As a young country, India has a lot to be optimistic as well as skeptical about the AI-future. The expected decline of many of the current job roles and the creation of new ones will be a significant challenge for young Indians. Despite the widespread consensus of AI stealing jobs, industry leaders point that ultimately it would lead to a net increase in the number of jobs available. These new jobs will be of different types, just like different types of jobs that emerged after the first industrial revolution.

Additionally, a widening gap might unfold at the level of individual workers. Demand for jobs will shift away from repetitive tasks towards those that require creativity, empathy, and other social skills along with digital abilities. Job profiles characterised by repetitive activities or that require a low level of digital skills could experience the most significant decline as a share of total employment to around 30 per cent by 2030, from the current roughly 40 per cent.

According to McKinsey’s research, countries fall into four groups in terms of AI readiness. Currently, India is in the third group, which is described as economies which have a moderate ability to capture economic benefits from AI. “However, the benefits can be maximised, and losses can be minimised by putting necessary infrastructure and policy in place,” writes Senior Director at MeitY, Dr Sunil Kumar Srivastava in a research paper.

India’s AI journey so far has been optimistic despite challenges it faces as an emerging economy. It is nothing short of exciting to see what the country has achieved with AI so far despite the limited resources. With a thriving startup ecosystem, a young population and multinational tech companies, AI can be India’s black gold. And the key to unlocking that potential lies in preparing its students, workforce, policymakers, and entrepreneurs to overcome the challenges and to leverage the opportunities AI brings.

About the author

Jibu Elias

Content and Research Lead at INDIAai

Jibu Elias is the Content and Research Lead at INDIAai, currently lending his wide knowledge and keen insight into artificial intelligence for building a unified AI ecosystem in India. He specialises in ethical and legal implications of AI. He is an alumnus of the London School of Economics (LSE), where he studied International Relations, specialising in Sino-India relations.

Previous Article

Healthcare data collection during COVID19, Aarogya Setu and the dawn of a new era in data privacy. Interview with Dhruv Suri, Partner, PSA

Next Article

#IndiaAI360 webinar series: AI skills and what colleges are not focusing on

Want to get your article featured?

Leave your email address here so our team can contact you.

Suggested Articles