What SNU Robotics Institute can Learn from the MIT Media Lab.
Dr. Ali Shtarbanov received his PhD from the MIT Media Lab and joined the SNU Robotics Institute in September as its first postdoctoral fellow. His doctoral research in Human-Computer Interaction focused on designing hardware development platforms and deploying technological ecosystems that foster creativity, innovation, and interdisciplinary collaboration. He has also worked at Meta Reality Labs and IBM Research. During his PhD, he developed a framework for creating highly impactful, general-purpose platform technologies and maximizing their impact by building ecosystems around them. He developed multiple hardware and software platforms, including FlowIO, SoftRobotics.IO, SleeveIO, and ModiStrap, along with dozens of application-oriented projects enabled by them. Among these, FlowIO, together with the SoftRobotics.IO ecosystem built around it, has grown into a widely recognized platform that has been used by researchers, designers, and engineers at more than 100 universities and institutions around the world, including Stanford, Harvard, Cornell, ETH Zurich, Imperial College London, NUS, NTU, and Tsinghua University, as well as at companies such as Meta Reality Labs, Toyota Research Institute, and several startups. The projects enabled by his work have spanned numerous fields and domains, including soft robotics, haptics, interactive interfaces, assistive technology, human-robot interaction, automotive applications, as well as education, art, and design. His work has enabled multiple master’s theses, art exhibitions, and dozens of published research projects across these domains, while giving Dr. Shtarbanov the opportunity to collaborate with more than 150 people from diverse backgrounds and institutions. Among the recognition he has received for his work are the Hackaday Grand Prize, the world’s largest open-source engineering competition; the Grand Prize at the Create the Future Design Contest, organized by NASA’s Tech Briefs magazine; the top prize at the ACM CHI Student Research Competition; and multiple design awards, including an iF Design Award. Most recently, he received the Brain Pool Fellowship, Korea’s premier fellowship for outstanding overseas scientists, through a competitive research proposal process administered by the Ministry of Science and ICT and the National Research Foundation of Korea.

Dr. Shtarbanov brings a unique perspective to the SNU Robotics Institute, combining the highly impactful, interdisciplinary research culture of the MIT Media Lab with his own experience developing platforms and ecosystems that have enabled technologies to spread across fields, institutions, and applications. He has a concrete vision for applying this combined approach to help shape the SNU Robotics Institute into a globally impactful, interdisciplinary institution that goes beyond developing great robotic technologies and publishing in top journals to developing and deploying entire ecosystems that enable these technologies to reach and be extended by thousands of people across diverse fields, while fostering highly interdisciplinary international collaborations.

Dr. Shtarbanov first learned about SNU when he met Professor Kyujin Cho at an IROS soft robotics workshop in 2023, where they were both giving invited talks. He was deeply inspired by the diversity of robotics projects Professor Cho presented and immediately began imagining numerous new applications for the technologies.

“I remember thinking about all the new applications those technologies could be used for, especially if they were extended into platforms and made more widely accessible for people to obtain, replicate, or extend. I saw an entire space of possibilities for how those robotics projects could be used in other domains, such as HCI, interactive interfaces, and art.”

“Many of my colleagues from other fields would absolutely love to know about these projects and use them in their own creative work. Why are they not more widely known outside of the domain of robotics? We could do so much with them and extend them in so many different directions.”

After the workshop, Professor Cho invited him to visit his lab at SNU and see the projects in person. Two years later, in December 2025, Dr. Shtarbanov finally had the opportunity to visit Korea and SNU while doing an international speaking tour. He was so delighted by the experience that he extended his stay in Korea to a full month.

“The research was incredibly inspiring, the people were brilliant, and I received a very warm welcome from everyone I met at SNU. I was also incredibly impressed by the facilities and equipment available, which matched or even exceeded those I had access to at MIT.”

He also enjoyed SNU’s scenic campus and many aspects of living in Korea. More importantly, he envisioned numerous ways to collaborate with SNU researchers: deploying existing projects, extending them into platforms, and building entire ecosystems around them.

“I see enormous potential in the technologies developed at SNU to reach a far wider and broader impact beyond publications, outside the domain of robotics, and across fields—and I want to work with people at SNU to unlock that impact potential.”

“Impact”—a word that initially puzzled people at SNU
Dr. Shtarbanov used the word “impact” far more often than “research.” Impact is hardly an unfamiliar term to researchers at SNU. If anything, it is a word they are required to include in almost every research proposal. Researchers often struggle to imagine and describe “impact” before their work has even begun, as though the research were about to lead immediately to commercialization or a startup.

For people accustomed to writing those ambitious projections, his insistence that researchers should think more deeply about impact was, frankly, somewhat puzzling.

But the impact he describes is not a grand promise written in advance of the research. Nor is he saying that every paper must immediately lead to technology transfer or a startup. Those outcomes, he explains, happen in stages. To him, impact is closer to the work that begins after the paper is published.

So what does he mean by impact, particularly in the context of a technological project?

Broadly speaking, he says there are at least three factors by which he defines the true impact a technology project. The first is the number of users or beneficiaries served by the project. The second is the number of use cases or utilities the project offers. A technology that serves five different needs, for example, has greater impact potential than a technology that serves only one, all else being equal.

The third factor is the number of progenies it enables. If a project engenders new discoveries and new projects, it creates what he calls a “posterity impact.” Its impact includes not only those immediate progenies, but also the future generations of work that they enable.

According to his philosophy of impact, the most impactful projects are therefore those that maximize all three components: beneficiaries, utilities, and progenies. The examples he gives include building blocks like the transistors, or tools and platforms like Raspberry PI, Arduino, and Scratch.

Research that Creates Ecosystems not just Publications
At first, it is easy to interpret his argument as a call to communicate good research more actively. Indeed, at the MIT Media Lab, researchers often build entire websites around many projects and give talks across many different venues, both academic and non-acadmic ones.

This is not about publicity, showing off, or building an individual researcher’s reputation among the general public. If a new piece of knowledge has been recognized through publication, researchers who do not read that particular journal may still need to know about it.

If the project is a robot, a materials scientist might see it and try a new material. A designer might create a different form. An artist or educator might discover an application that the original researcher had never considered. Research has to reach people outside the journal’s immediate readership before connections across entirely different fields can begin.

But the impact Dr. Shtarbanov describes goes beyond communication. Publishing a project website and giving talks are not enough. For research to be used and extended by others, communities, deployment models, infrastructure, and supporting technologies must also be developed around it. This is what he calls an “ecosystem.”

“To maximize the impact potential of one’s work, it is important to look beyond publications. Instead of taking a publication-centric approach, one should take an ecosystem-centric approach. The new technology that is developed is only one component of an ecosystem.”

Why did Scratch and LEGO Mindstorms become better known than their papers?
Dr. Shtarbanov points to Scratch and LEGO Mindstorms as two widely-known projects associated with the MIT Media Lab. Scratch, a programming environment designed for children, is now widely used to introduce coding to elementary school students, including in Korea. Almost everyone interested in technology knows Scratch, yet relatively few people have read the academic publications behind it. The same applies to LEGO Mindstorms.

“If we consider some of the most well-known Media Lab projects, such as Scratch or LEGO Mindstorms, they did not become impactful because they were published in a top journal. You have probably never even heard of any of the publications about those projects. And it was not just because of publicity either. Rather, it was because entire ecosystems were built, developed, and deployed around those works.”

FlowIO was also designed not merely as a technology described in a single paper, but as a platform. It combines miniature pneumatic hardware, sensors, and software in a modular system that enables not only researchers, but also students, artists, and makers, to develop soft robotics projects more easily.

Dr. Shtarbanov also founded SoftRobotics.IO, an online community where people can share projects, propose new modules and software, find collaborators, and extend one another’s work. For him, building this community and enabling other people to create with the technology are themselves part of the research.

Turning an idea into a technically sophisticated paper is difficult. But unpacking the highly developed idea contained in that paper and turning it back into a form that other people can understand, obtain, deploy, and extend is far more demanding and rewarding.

According to Dr. Shtarbanov, the MIT Media Lab takes a very different approach to research from many other departments.

“Others typically take a top-down approach to research, where they have an objective to publish first and then develop their project for the sake of publication. At the Media Lab, we often start by just building something first, without much consideration of what field it would fall into – science, engineering, medicine, art, etc. And only when we have something working, we then start asking the question how to disseminate it to the world, what field it would fit into, where it would make sense to publish it, and what other forms of dissemination would also be appropriate – exhibition, website, app, a spin-off, a non-profit entity, TED talk, press-interview, design-competitions, op-ed, etc.

Of course, the MIT Media Lab is unconventional even by American university standards. Simply transferring its research practices to SNU would not automatically produce the same results.

What he seeks to bring from the MIT Media Lab to SNU Robotics

Those expecting advanced new equipment or a large research initiative may be surprised by his answer: “Reogranize the space so it itself fosters more interaction and communication between people; make demos and informal gatherings part of everyday life; host social activities within the lab spaces themselves to enable researchers from different labs to informally meet, talk, and visit each other’s labs; build ecosystems around prominent projects; and use English for all official written and spoken communication within the Institute. These are all parts of the DNA of the MIT Media Lab, and we can replicate them here at SNU RI as well.”

He suggests starting with something as simple as creating a social area in the most prominent part of the research space, so that people naturally encounter one another while going about their work. He sees this not simply as a social event, but as an initiative with important strategic value: one that builds community and cohesion and can spark new creative ideas. Another Media Lab tradition he hopes to recreate is a scavenger hunt game for incoming researchers. Demos, he argues, should also become part of the mundane, daily experience. They should feel natural and spontaneous rather than being treated as formal presentations or something reserved for special events. At the Media Lab, there is a “demo-and-deploy culture.”

Declaring a commitment to interdisciplinary research or formally placing researchers from different fields in the same funded project does not necessarily make meaningful collaboration happen. New ideas often emerge only after people repeatedly encounter one another, see what others are working on, and exchange informal questions.

For Dr. Shtarbanov, sofas, tea hours, and everyday demos are not simply amenities or social activities. They are part of the infrastructure required to build a research ecosystem.

Five-Part Plan of Action by Dr. Shtarbanov

If the SNU Robotics Institute is to realize its ambition of becoming a globally influential research institute, I believe there are five things that can be done immediately. None requires the Institute to abandon its existing strengths. Rather, they are relatively simple changes that can make those strengths more visible, connected, and impactful.

1. Make English the official language of communication

All official spoken and written communication within the Institute should be conducted in English. This includes meetings, shared documents, calendars, emails, announcements, articles, and especially group chats.

For an institute that wants to attract and retain international researchers, this is not simply a matter of convenience. It is basic infrastructure. When important information is communicated only in Korean, international researchers are excluded from conversations, decisions, and informal exchanges that their Korean colleagues take for granted. It directly reduces productivity and makes it much harder for foreign researchers to become fully integrated into the research community.

This is arguably the simplest and most important change the Institute can make toward becoming genuinely international. It does not require a new building, new equipment, or a new funding program. It simply requires establishing a clear institutional norm and consistently following it.

2. Design the physical environment to encourage interaction

The physical layout of the Institute should actively encourage people to encounter one another.

A prominent social area—with sofas, a coffee table, board games, and perhaps a foosball table—should be added to the research space, and be in the most prominent area of the space, rather than hidden in a separate lounge. People should be able to see the space from their desks and immediately recognize when others are there. This creates opportunities for spontaneous conversations that otherwise would never happen.

The space should also be designed as a point of attraction for researchers from other laboratories. A visible activity such as foosball can serve a strategic purpose: someone passing through the Institute might stop to play, meet researchers, see a demonstration, and discover work that they would otherwise never encounter.

The demos and social area should therefore become focal points of the Institute. Both provide natural destinations for people taking breaks, and both create opportunities for researchers to encounter new projects and new people outside their immediate circles.

The goal is not to make the Institute more comfortable for its own sake. It is to use the physical environment as social infrastructure for research.

3. Create recurring mechanisms for interdisciplinary interaction

Simply putting researchers from different fields in the same building does not guarantee interdisciplinary collaboration. People need mechanisms that repeatedly bring them into contact with one another.

I propose starting with several simple initiatives:

1. Scavenger Hunt Project Finding Game— A tradition we have at the Media Lab is for incoming students to do a scavenger hunt game on their first day, where they have to walk around all the labs and find as many projects as they can from a list of 50 projects during an afternoon, talk to the researchers that made them, and get a sticker from the corresponding researcher. Whoever collects the most stickers wins and the tops 3 winners get prizes. This game has an important strategic function that right from day one, incoming students become familiar with dozens of researchers at the lab, get to talk to them, and become familiar with their work. It build social cohesion. This also provides a structured way for newcomers to discover the expertise already around them.
2. Normalize Demoing — Make showing unfinished work to colleagues a routine part of everyday life rather than something reserved for formal presentations or special demo days. The Media Lab has a demo-and-deploy culture that allows people to encounter ideas early, provide feedback, and discover unexpected connections.
3. Weekly or Biweekly Tea + Snacks — Hold a casual tea-and-snacks hour directly inside the Institute’s research space, and invite researchers from other laboratories as well. Different groups can take turns hosting it. The objective is not simply socialization, but to create a recurring opportunity for researchers who would otherwise never meet to encounter one another and do so in a space where they are surrounded by demos which can spark new ideas.
4. Cross-Lab Project Exchanges — Regularly identify projects, technologies, or research questions that could benefit from expertise elsewhere in the Institute and actively connect the people involved. The Institute should make it easy for a researcher to say, “I saw this project and I think it could be useful for my work.”
5. Interdisciplinary Free Lunches: People can sign-up to have lunch with someone from another lab in order to foster interdisciplinary interactions. And if they complete the lunch, they get reimbursed for it, up to some limit per year. This fosters community across field boundaries and reimbursement incentivizes people to do it.

These initiatives are part of the secret-sauce of the Media Lab, and they are quite simple to implement in the SNU RI as well. Interdisciplinary research often begins not with a formal collaboration agreement or proposal, but with a conversation between two people who happened to discover that they were interested in the same problem.

4. Adopt an ecosystem-centric rather than publication-centric approach to research

The Institute should measure the success of its research not only by what gets published, but by what happens after publication.

A publication should be treated as the beginning of dissemination and deployment, not the end of a project. Strong projects should be communicated through project websites, exhibitions, talks, educational materials, open-source repositories, a multidude of applications, spin-offs, replication instructions, social media, press, and collaborations with organizations and people outside academia.

This also means deliberately reaching beyond the field of robotics. A robotics project may have enormous value to an HCI researcher, designer, artist, educator, materials scientist, or entrepreneur who would never encounter the work through a robotics conference or journal.

And Korea offers an especially unusual opportunity. Seoul is home to a huge concentration of globally influential cultural figures and media organizations. The Institute should explore collaborations with Korean celebrities, creators, artists, and other public figures who could introduce technologies developed at SNU to audiences far beyond the traditional academic community.

But dissemination is only the first step.

Research projects should not be abandoned once their papers are published. This is often when the real work of creating impact begins.

Conversations with students have revealed that many past projects and demos become extremely difficult to replicate after the original researcher graduates. The tacit knowledge required to rebuild a system leaves with that person. As a result, when a new researcher wants to build upon an earlier project, they just start from scratch because tacit knowledge is not well preserved.

The Institute should identify its most valuable projects and deliberately build ecosystems around them. This entails documenting them thoroughly, making hardware and software accessible, creating replication instructions, maintaining online infrastructure, developing modular extensions, supporting users, and encouraging communities of researchers, students, artists, and makers to contribute new applications.

The objective is to create a chain in which one project generates another project, which generates another, and so on. This is the “posterity impact” that allows the influence of a research project to continue growing long after its original creators have moved on.

5. Bring in top talent from diverse fields and with diverse international perspectives

Finally, the Institute should broaden the kinds of people it attracts.

Robotics naturally requires excellent engineers, but the most interesting future applications of robotics will increasingly emerge at the boundaries between engineering and other fields: HCI, design, art, materials science, medicine, education, architecture, fashion, and many others.

The Institute should therefore actively recruit outstanding people who do not fit neatly into the traditional robotics profile, including researchers and practitioners from different disciplines, countries, and research cultures. And this is likely the easiest part of this plan.

If the first four pieces are successfully put into place, the talent pool would naturally become larger and more diverse. An institute that communicates in English, provides an unusually collaborative environment, makes its research accessible to the world, builds ecosystems around its projects, and develops a reputation for actually turning research into impact will become a much more attractive destination for exceptional people from around the world.

The five parts are therefore not independent initiatives. They reinforce one another:

Communication enables participation.
The physical environment enables encounters.
Interactions enable collaboration.
Ecosystems enable impact.
And impact attracts talent.

Together, they can help transform the SNU Robotics Institute from a collection of excellent laboratories into a genuinely connected research ecosystem—one in which ideas do not simply get published, but continue to move, evolve, and generate new ideas long after the original research is finished.