Seoul students win WRO award for robot-training data system

Three Seoul students won one of four top prizes in the World Robot Olympiad STEM Entrepreneur Program after building PlayData, a system that turns everyday human movement into training data for robots and Physical AI. The team earned $5,000 and six months of mentoring as WRO expands the program beyond its inaugural event in Hyderabad.

Why it matters: - Robots and Physical AI systems need more real-world interaction data to learn how humans grasp, twist, pull, insert and press objects. - The Seoul team’s project points to a lower-cost way to capture that data from everyday human activity instead of paid demos. - The WRO STEM Entrepreneur Program is designed to push student robotics projects toward market use, not just competition judging.

What happened: - Three students from Seoul were chosen as one of four winning teams in the World Robot Olympiad STEM Entrepreneur Program. - The team competed as Why Combinator. - The students are Jaeseong Kim, 16; Sungchan Kim, 17; and Sungkyeom Kim, 16. - Their project, PlayData, turns everyday human interactions into training data for robots and Physical AI. - The team presented the project in Hyderabad, India, during the WRO STEM Entrepreneur Program on September 28, 2026. - The event brought together 22 student teams from 13 countries.

The details: - Each student came to the same problem from a different robotics project. - Sungkyeom had been building robot hands but lacked enough manipulation data to train them effectively. - Sungchan had been teaching robots physical tasks and also lacked enough examples for training. - Jaeseong faced similar data shortages while researching humanoid robotics. - The students realized they were all confronting the same gap in robot training data. - PlayData looks at activities people already do, including escape rooms, as a source of physical interaction data. - The team built a data glove and body-camera prototype. - The prototype records first-person video, finger movement, hand motion and contact force. - The students also built a robotic hand with 16 independently controlled movements. - The team created a data-processing system that synchronizes and structures the recordings. - The WRO STEM Entrepreneur Program is powered by Aramco. - The program asks students to explain how their technology works, why it matters and how it could move from competition to market. - WRO said the Hyderabad event was part of its Future Innovators competition pipeline. - Sang-Yep Nam, national organizer for WRO in Korea, said Why Combinator stood out for identifying a real challenge, combining their knowledge and building a concrete solution. - Four winning teams came from Korea, India, Italy and the Philippines. - Each winning team receives a $5,000 award. - Each winning team also gets six months of mentoring from industry experts. - Mentors include investment professionals from Aramco Ventures and experts from India STEM Foundation and WRO.

Between the lines: - The project reflects a broader shift in robotics from hardware-only innovation to data collection and training infrastructure. - The WRO format is trying to teach students how to think like founders as well as engineers. - Aramco is using the partnership to link STEM education with entrepreneurship and real-world innovation. - Hadeel Al-Eisa, Aramco Corporate Citizenship Manager, said the company wants to help young people create, build and solve beyond the classroom. - For the students, the award also addresses a practical constraint: limited resources. - Jaeseong Kim said the support lets the team focus on hardware, research and moving the idea closer to the real world.

What's next: - WRO plans to bring the STEM Entrepreneur Program format to national organizers around the world. - The winning teams will spend the next six months developing their projects with mentor support. - The Seoul team will continue refining PlayData for use in the Physical AI industry.

The bottom line: - A student robotics team from Seoul turned a common research bottleneck into a prize-winning product concept, showing how human play could become a useful source of robot-training data.