Every year, university researchers make discoveries that could change the way we treat disease, power our communities, or improve the technologies we rely on every day. But promising discoveries don’t always make it out of the lab.
A researcher may have years of validated data behind an invention and still have no clear answer to the questions that determine whether it reaches the market: Who would buy it? How large is the opportunity? Is the best path forward a startup, a licensing agreement, or an industry partnership?
This “commercialization gap” is one of the biggest challenges in academic innovation. While researchers are experts in advancing scientific knowledge, evaluating a technology’s commercial potential requires a different set of skills. The Mack Institute has long supported Penn researchers in this process through programs like the Y-Prize Competition and the Penn-Wharton Commercialization Workshop.
The Mack Institute’s newest initiative in this space is Commercialization of Academic Science (CAS), an experiential learning course that pairs small teams of Wharton MBA students with Penn researchers working to bring early-stage technologies closer to market. Over the course of a semester, students help researchers tackle commercialization questions through customer discovery, market analysis, competitive research, and business strategy. In return, they gain hands-on experience tackling real commercialization challenges at the intersection of science, innovation, and entrepreneurship.
“Students are handed a real technology with no established path forward, and they work alongside a scientist who may be deciding whether to devote the next decade of their career to it.”
In the past year, CAS has supported 14 projects with 50 students across an intentionally broad range of disciplines. Projects have explored everything from new dental hygiene products to AI-enabled cancer treatment and 3D imaging for autonomous systems. The class is taught by Aswin Mannepalli, a Mack Institute Senior Fellow and co-founder of the customer discovery platform Custard, who created the course with the Mack Institute’s Valentina Goutorova.
“The range of projects is deliberate,” he said. “Innovation at Penn does not sit in one department, and neither should the course. In the same semester, one team may work on a consumer product while another tackles a technology headed toward an FDA pathway.”
For example, one team partnered with Dr. Rafe McBeth, Assistant Professor of Radiation Oncology and Director of AI at Penn Medicine, to help commercialize a technology that uses AI to streamline radiation treatment planning. McBeth’s lab had already developed a proof of concept: a chain of AI models that compressed treatment preparation from nine days to under two hours. But the researchers weren’t sure how to position the technology. Rather than framing it as another AI model, the CAS students encouraged them to think more broadly about the workflow challenges it solved.
“The students told us that the opportunity is not just better AI models,” McBeth said. “It is a workflow orchestration engine that coordinates handoffs, minimizes idle time, and embeds itself deeply into clinical workflow.”
Colin Twomey, Executive Director of the Data Driven Discovery Initiative in Penn’s School of Arts & Sciences, has worked with CAS for multiple semesters. He created UAtlas, an interactive map of Penn’s research ecosystem that allows users to explore publications by topic, school, department, author, and time period. Originally designed to help Penn faculty and staff discover one another’s work, Twomey soon realized the platform could have applications at other research universities.
To explore that potential, he turned to CAS. Students helped him develop a better user experience and a plan for scaling the project, complete with feedback from potential clients. In the Fall semester project, students conducted structured interviews with potential users, walking them through the platform to see what worked and what didn’t. These interviews flagged a missing feature: beyond exploring the map, users also wanted to be able to search in a targeted way. In other words, they wanted the “ChatGPT experience”: to type a question (“Who at Penn is working on CAR-T cell therapy?”) and get a precise answer.
The spring team picked up where the fall team left off. With a clearer picture of the user base in hand, they focused on the larger question: how does UAtlas become sustainable? They interviewed stakeholders at R1 universities beyond Penn—such as Georgia Tech and NYU—to test whether the platform’s value proposition would translate.
By the end of the project, Twomey had not only a better understanding of how users wanted to interact with UAtlas, but also a clearer commercialization strategy for taking the platform to other institutions. Reflecting on his experience, Twomey said that working with Wharton MBA students challenged him to think about his research in a different way.
“With commercialization, there’s a language involved: a different mindset or way of thinking,” he said. “Having access to someone who already has that language and that mindset—and is able to translate your thinking into a way that’s actionable in that other world—is incredibly valuable. You learn a lot from going through that process.”
“With commercialization, it’s not just what’s possible, but what’s sustainable? Is there a sustainable way to make this exist and provide a service for others?”
The experience also clarified what he sees as the fundamental difference between research and commercialization.
“Research and commercialization are different types of problems,” he said. “In research, you’re being led by the frontier of what’s possible: you figure something out, and that leads to the next hypothesis. With commercialization, it’s not just what’s possible, but what’s sustainable? Is there a sustainable way to make this exist and provide a service for others? That service mindset—what does it take to build something that keeps delivering value over time? — is very different than the research mindset.”
For Mannepalli, inspiring that kind of clarity is the real measure of success for the course.
“People assume the goal is to turn every project into a startup. It isn’t,” he said. “Often, real success is helping a researcher make the next decision with better evidence than they had at the start of the semester. Students are handed a real technology with no established path forward, and they work alongside a scientist who may be deciding whether to devote the next decade of their career to it.”
“You cannot manufacture those stakes,” he continued. “Penn has dozens of projects at exactly that moment. Our job is to open the door.”