IBM – Subram Natarajan sees AI use in vaccine supply chain management, predict new waves
IBM, which is working on many Covid-19 projects around the world, is convinced the pandemic has triggered an acceleration in innovation and adoption of technologies like artificial intelligence, with new use cases in retail and banking, but more importantly in managing and now mitigating a global pandemic.
Subram Natarajan, Chief Technology Officer of IBM India, has seen “dramatic changes” this year, along with a “fair bit of confluence of various technologies, AI being one of them”. While the lockdown forced many businesses to go digital to survive and adopt technologies like AI, Natarajan says it was the effectiveness of artificial intelligence in citizen’s services that have had the most significant impact.
Natarajan cites the Indian Council for Medical Research’s (ICMR) collaboration with IBM to implement the Watson Assistant (an AI-powered query-answering system) on its portal to respond to specific queries on Covid-19 from frontline staff and data entry operators across testing and diagnostic facilities across India. “They (ICMR) had a closed set of health professionals, whether it is the front end, people who are administering the tests, doing the analysis, the support staff, or the core health professionals at the backend. They were trying to really figure out how to get the unified message across to the wider population in India,” he remembers a time when “there were more questions than answers”.
“The questions were coming from pretty much all angles, whether it is the general population or health professionals.”
The Andhra Pradesh National Health Mission too has a Watson virtual agent that disseminates Covid-19 related information in English, Telugu and Hindi on the response efforts and measures by the Department of Health, Medical and Family Welfare. Globally, IBM Research has developed a cloud-based AI research service that has ingested thousands of papers from the Covid-19 Open Research Dataset (CORD-19) and licensed databases from DrugBank, Clinicaltrials.gov and GenBank, allowing users to make specific queries to extract critical COVID-19 knowledge.
“The adoption of AI has seen a gradual metamorphosis… What this pandemic has done is accelerated how we as individuals, as normal people, consume AI,” explains Natarajan who has been with IBM for over two decades.
“Banking services, which was already in the verge of getting sort of digitised to some degree, has all of a sudden graduated into more than 90 per cent of all transactions being handled by digital,” he says. Even from the consumer side, it has accelerated acceptance of AI and other digital technologies. “I don’t think we would go back to a day when things are much more human driven… We would expect human involvement and engagement in a much more valued part of the workflow,” he explains, adding how users have also started expecting bots to respond to the more mundane queries.
With large companies, the acceptance, however, is not that easy. But Natarajan says it is “no longer a choice”. He elaborates on the example of Parle’s supply chain which used to be governed by conventional wisdom, more statistics driven with a straight away projection of how much goods would be consumed. “But with the pandemic, nobody was able to predict how much inventory one should hold primarily because the consumption pattern was completely disrupted. There were lot more factors at play,” he says. This is where IBM’s Watson-powered ‘intelligent supply chain’ solution came into the picture. “Once these additional parameters regarding Covid-19 were fed in, we were able to predict demand in a more intelligent manner, helping Parle reduce time to market and right size the inventory for its perishable goods.
This AI-drive supply chain management will also be in play when IBM helps governments rollout to the vaccines to millions of citizens. “The idea would be to clearly understand the different patterns of the pandemic, where the density is, where the vulnerable are, where the capability to distribute the vaccine is very high. So there are hundreds of data characteristics or parameters. Once you feed in a well-defined model of what is the most efficient way of distributing the vaccine… a prescriptive way of handling the vaccine that can be provided, and the AI plays a vital role in this whole distribution,” explains Natarajan.
IBM has made available an open vaccine-management platform to facilitate the delivery bringing together the power of blockchain, data and AI, security, and hybrid cloud. There is also the high performance compute consortium — a global initiative by IBM in collaboration with the US Department of Energy— created just after the pandemic surfaced to make hundreds of petabytes of computing capacity available to the scientific population around the world, particularly those in the health care and life sciences space. NIT Warangal was recently approved to run an experiment as part of the same.
“AI is like a child, you start teaching it and it slowly becomes much more knowledgeable. We have not had a lot of simulations of this scale (Covid-19 pandemic) done before,” Natarajan says, adding that however, one of the learnings this year has been that an occurrence in one particular geography can be repeated elsewhere. “A good example is the second wave. A lot of countries went through this… some countries kind of lagged behind, others were ahead of the curve,” he says, adding how this learning can be used to predict and prepare for a second wave elsewhere.
Natarajan emphasises that all this has shown AI is no longer an option. “It is a question of how and when one should get on board.”
