23 Jun 2021
News
7 minute read

New IBM-IISc innovation lab: How hybrid cloud and AI could transform India

The new IBM-IISc Hybrid Cloud Lab will be housed on the IISc campus in Bengaluru. The aim is to get scientists and students to solve problems in information management, develop more efficient verification and optimization technologies, and establish better testing frameworks to write code—to further advance the emerging hybrid cloud for AI research.

The India Institute of Science campus (credit: IISc)

Twelve Nobel Prizes. A booming tech scene with established giants and some 100,000 startups dotted across Delhi, Bengaluru, Mumbai, Hyderabad, Pune and Chennai.

A research institute that has just been ranked as the world’s top research university. An ambitious space program with probes sent to Mars and the Moon. And colors, colors everywhere—from deep orange marigold flowers to different hues of red, purple, yellow on clothing, buildings, markets, as far as the eye can see.

This is India. And the institute that was given a perfect 100 out of 100 score by the QS World University Rankings 2022 is the Indian Institute of Science (IISc). The same institute that is now partnering with IBM Research to create a new innovation lab in Bengaluru—the IBM-IISc Hybrid Cloud Lab.

The aim is to further advance the research in artificial intelligence and hybrid cloud in this already very tech-savvy nation.

IBM Research has been in India since 1998, continuously producing crucial applications across the country. From retail to healthcare and agriculture, the scientists working at IBM India keep pushing the limits of possible, mainly in artificial intelligence.

“With deep multidisciplinary skills across AI, software engineering and cloud computing, IBM’s research lab in India is already at the forefront of our work in AI for IT. It has produced innovations that have contributed to app and data modernization, AIOps, compliance, and so much more,” says Sriram Raghavan, Vice President of IBM Research in AI.

AI, yes, but also hybrid cloud—the world’s largest virtual computer. Enshrouding the globe thanks to hundreds of data centers, hybrid cloud is critical for further advances and adoption of AI across nearly all industries.

“We envision a world where the transformation is led by a movement to the cloud, making applications a lot more agile—and able to drive innovation faster,” says Gargi Dasgupta, the Director of IBM Research India.

“Cloud is almost like our first principle—and after you’ve moved your applications to the cloud, you can continuously infuse innovations into your line of business.”

With the new partnership, IISc and IBM are getting ready to conquer new heights in cloud and AI.

Bringing academia and industry together

Housed on the IISc campus in Bengaluru, the lab will unite scientists from two different, yet so similar, worlds, academia and industry. Different in structure perhaps, but similar in the amount of talent, creativity and scientific rigor.

After all, says Yogesh Simmhan, associate professor at IISc’s department of computational and data sciences and co-chair of the new lab with Amith Singhee of IBM Research India, advancing hybrid cloud technology is not trivial. It requires a robust understanding of the fundamental problems as synthesized from real-world applications and users, and deep knowledge of complex data, computing and AI and machine learning systems.

“Industry and academic researchers bring together synergy on each of these parts,” says Simmhan, “and help define meaningful and hard problems and innovative solutions to them to achieve high-impact outcomes.”

It should also help students develop the necessary skills in AI for Hybrid Cloud—an emerging, exciting field.

Dasgupta is certain that the lab will lead to innovation, collaboration and conversations in the broader technical community in the region community through publications, workshops and open-source software contributions.

“We will be conducting research in the open, do joint presentations of our findings at conferences and in journals, and release open-source material to the community. Open source accelerates innovation at a much faster rate and scale and helps communities to develop skills and drive the results of research faster into the ecosystem,” she says.

The center will aim to solve problems in information management and better verification technologies, and build autonomous, self-healing computing systems. And it will help develop AI systems able to code and create more efficient optimization technologies.

In India, just like elsewhere in the world, advances in AI can help a lot. From autonomous IT operations to future self-driving vehicles—all these applications are on the cloud. A single minute of downtime of an online app means losing a lot of money to the business. With the cloud, the possibility of power outages is shrunk to the minimum.

And cloud pulls together different computer science disciplines, says Dasgupta—such as distributed systems, serverless computing, information management, data and AI. “That’s why we needed a top science institute to work with us. And I’m sure that together we’ll solve many fascinating problems.”

Bird's eye view of outskirts of Bengaluru, from top of Nandi Hills.
Figure 1:
Bird's eye view of outskirts of Bengaluru, from top of Nandi Hills.

Future communities of discovery

IBM Research has been pushing the idea of public-private partnerships for years. After all, talent knows no labels. Be it a clever physicist, chemist, mathematician or engineer from a university lab or a just as clever researcher from a research department at a corporation, together they are much more likely to bring a brighter future a tad closer—faster.

IBM’s global AI Horizon Network is a notable example. In India, IIT Delhi and IIT Bombay joined the network in 2018. Both partnerships have already made significant progress in NLP and AI—from question-answering to multimodal and neurosymbolic AI.

One outcome of the joint effort with IIT Bombay has been enabling Watson AI to understand the Hindi language in Devanagari, the writing system for Hindi—including sentence structure and grammar. Researchers with the AI Horizon Network have jointly published a lot of high-quality papers at top conferences for machine learning and AI, and the network has enabled rich experiences for students through internships and mentorships from academia and industry.

Another example of a successful public-private collaboration, outside India, is the MIT-IBM Watson AI Lab in the US. Researchers from both organizations write joint proposals that are evaluated by reviewers from both MIT and IBM. The scientists then execute them jointly. Established in 2017 for a decade-long partnership, the lab has had over 50 approved projects and hundreds of publications, more than any other industry-academia collaboration in the world.

It's yet more proof that bringing brilliant minds from industry and academia together is a recipe for success, a sure way to push the boundaries with future-looking independent thinking, talent and determination.

Times are changing. Academia’s involvement with industry should no longer be about a university asking a company to write a check. The university shouldn’t be doing research and development on its own, either taking the product open source or licensing it. Instead, academia-industry partnerships should be about accepting that scientific and technological talent is distributed. That we can dramatically speed up innovation by bringing together funding, infrastructure and talent—to jointly design and execute research projects.

We can’t afford to procrastinate. With more government and private funding and more strategic partnerships between industry and academia like the new India collaboration, we can and should boost innovation. Only together we will halt future crises and help save the world.

We have to act now.

Date

23 Jun 2021

Authors

Tags

Share