Visiting professor Satish Narayana Srirama on the Mobile & Cloud Lab and practical IT education

Satish Narayana Srirama worked at the University of Tartu for 11 years, from 2009 to 2020, first as a postdoctoral researcher and later as a lead researcher in big data. He is currently a professor at the University of Hyderabad in India, but remains connected to the University of Tartu as a visiting professor and as the founder and member of the Mobile & Cloud Lab.

Our interview with him unexpectedly became so long that we split it into two parts. In the first part, Satish talked about his journey from a small village in southern India to Germany and then to Estonia. In this second part, he discusses the Mobile & Cloud Lab, cloud, edge, and fog computing, teaching, and what future researchers should learn in the age of artificial intelligence.

How did the Mobile & Cloud Lab begin?

When I joined the University of Tartu, I was part of the Distributed Systems group and worked on scientific computing on the cloud. With motivated students such as Pelle Jakovits, Vladimir Sor, and Toomas Römer, we studied large-scale data analytics on the cloud and the migration of enterprise applications to cloud environments.

Then I combined that cloud computing work with my doctoral research on mobile hosts in enterprise service integration. That gave us a new research direction: mobile cloud.

A few more highly motivated students and researchers joined, including Huber Flores, Carlos Paniagua, and my first PhD student and postdoctoral researcher Chii Chang. As the research output became more visible, we named the group the Mobile and Cloud Computing Laboratory, or Mobile & Cloud Lab.

A major turning point came with support from a PUT startup grant in 2014, which enabled the lab to consolidate its work, attract more talent, and establish itself as a focused research group. At its height, the group included a professor, a lecturer, three postdoctoral researchers, five PhD students, and several master's and bachelor's students.

What are you most proud of from your years at the Institute of Computer Science?

I am proud that my growth in Estonia was not one-dimensional. Academically, we published in strong journals and conferences in cloud computing, IoT, distributed systems, and related areas. My work in cloud computing also received global recognition, and I had the opportunity to co-lead the Cloud Computing Manifesto with Prof. Rajkumar Buyya from the University of Melbourne.

But I am equally proud of the students. Several of them went on to build successful companies or contribute strongly to industry. Examples include Vladimir Sor's Plumbr, which originated from his PhD research, Carlos Paniagua's Glia, and Toomas Römer's ZeroTurnaround.

We also established an Internet of Things Lab with support from Telia. It included smart home and smart office demonstrators and attracted a wide range of visitors, from ambassadors and rectors to schoolchildren and teachers.

How did students shape the lab?

I could not have achieved these things without the quality of students I had at the University of Tartu. Several students stayed with the Mobile & Cloud Lab from the day they joined until they completed their studies or settled into their academic careers. Jakob Mass, for example, joined me as a bachelor's student and continued through his master's and PhD in computer science.

In the lab, we did not maintain much hierarchy. We believed in collective growth, healthy discussion, and competition. During the early years, Marlon used to joke that we were a "band of brothers."

The students also helped keep the research practical. Many students in Estonia were working in companies or running their own companies. They often brought real problems from industry to the lab, and we tried to offer better solutions from an academic perspective. This is one reason Estonia's startup ecosystem has been so successful.

How has your research evolved over the years?

My PhD work, from 2004 to 2008, was in mobile web services. The basic idea was to access and provide web services from smartphones. My postdoctoral work from 2009 focused on scientific computing on the cloud: large-scale data processing with MapReduce, migration of applications to the cloud, and cloud economic and deployment models.

Around 2011, we combined cloud computing and mobile computing into a mobile cloud. We studied how mobile phones could invoke cloud services for tasks such as face recognition or sensor data analytics, where the phone itself did not have enough computational power.

When IoT gained major academic attention around 2015, our previous expertise became highly relevant. We moved into edge and fog computing, where computation is pushed closer to sensors and users in order to reduce latency, network load, and privacy risks.

Today my research focuses mainly on distributed edge analytics across the edge-fog-cloud continuum. We work on serverless data pipelines, virtual clusters of resource-constrained fog devices, federated learning, and the combination of fog computing with large language models.

What problems in edge and fog computing remain unsolved?

Fog computing has received significant research attention since around 2012. Researchers have worked on frameworks, simulators, resource management, placement strategies, quality of service, and economic models. But after more than a decade, we still do not see large-scale public or private fog networks that can support broad IoT applications.

Most of the literature still relies on pilot studies, small testbeds, or simulators. One important reason is that fog computing has not yet presented a clear business case for companies and participating individuals. There are also not enough large-scale commercial or academic testbeds where prototypes can be evaluated realistically.

Another emerging question is quantum computing. If powerful quantum computers become widely available, quantum cloud computing may become important. If fog computing is cloud in proximity, we should already ask what quantum cloud computing might mean for edge and fog systems.

Your recent books have a strong practical teaching focus. Did that come from Tartu?

Very much so. At the University of Tartu, we placed strong emphasis on the practical adoption of technologies. Around 2010, one of the best decisions at the Institute of Computer Science was that every lecture should have a tutorial or lab component. Even theoretical courses were not exceptions.

When I moved to the University of Hyderabad, I wanted to bring that Tartu teaching style with me. I began adding weekly exercises to the courses I taught. Then I realized that there were not enough good textbooks to support this approach, so I started writing books with a strong hands-on component.

The first edition of Mastering Cloud Computing was published in 2013 and became widely adopted. The new edition incorporates developments from the last decade, including containers, cloud-centric IoT, fog computing, clouds for distributed AI and machine learning, DevOps, and quantum cloud. I worked on the second edition with Prof. Rajkumar Buyya and my former UT PhD student Shivananda Poojara.

The second book, Mastering DevOps: A Cloud Engineering and Data Science Perspective, grew out of the DevOps course developed and taught by Chinmaya Dehury at the Institute of Computer Science.

How does writing a book differ from writing research papers?

In research papers, the focus is novelty and presenting results properly, often within strict page limits. With textbooks, the focus is explanation. The goal is to help the reader understand the subject and gain confidence.

Books also require convincing a publisher that the work has value for readers and is viable for them. In that sense, I found writing and publishing books more difficult than publishing a strong research paper.

What makes a great master's or PhD student?

A great student is driven by curiosity and a genuine passion to understand and solve problems. Technical ability matters, but perseverance is just as important: the willingness to learn from failures and continue forward.

Good students take ownership of their work, think independently, and are not afraid to explore new directions. At the same time, they collaborate well, communicate clearly, and remain open to feedback.

What advice would you give to current UT students?

Make the most of the freedom and opportunities the University of Tartu offers. Be curious, explore beyond your coursework, and do not hesitate to take on challenging problems. Build strong fundamentals, but also develop the ability to think independently and question assumptions.

Research is incremental. Progress comes step by step, not in sudden leaps. Talk to professors, collaborate with peers, and seek feedback early and often. When you attend conferences, make a conscious effort to talk to people, discuss your research, and learn from others. These interactions often shape ideas and future collaborations.

Young researchers today also need critical thinking in the age of AI. AI can accelerate research and improve productivity, but there is also a danger of over-reliance, reduced depth of understanding, and trend-following instead of original ideas. Strong fundamentals, rigorous validation, and research integrity matter more than ever.

International mobility is also extremely valuable. It exposes you to different research cultures, ways of thinking, and global networks. Use mobility for learning, contributing, and bringing knowledge back to strengthen the research ecosystems around you.