We are delighted to share that the student team from the Intelligent Medical Computing Laboratory (IMCL, aka GEM) at Macao Polytechnic University has won third place in the Developer Open Group of the Physical AI track at BEYOND Hack Day 2026.
The award-winning team, made up of Jiefeng Xie, Shengxi Chen, and Jianchang Ruan from the Faculty of Applied Sciences, competed under the guidance of Yue Sun. During BEYOND Expo 2026, they took part in the BEYOND Hack Day competition and stood out through their strong performance in a highly demanding real-world robotics challenge.
This achievement reflects the team’s broad capabilities in artificial intelligence, computer vision, deep learning model training, and intelligent system development. It also highlights the ability of MPU students to turn research knowledge into practical engineering solutions in authentic application scenarios. Notably, the team was the only winning university team from Macao in this event.
1. Competition Achievement
BEYOND Hack Day 2026 was one of the major innovation activities held during BEYOND Expo, bringing together AI developers, university students, research teams, entrepreneurs, and innovation practitioners. The Physical AI track centered on real robotic tasks and attracted 40 participating teams across junior secondary, senior secondary, and open developer divisions.
In the Developer Open Group, the MPU team designed and implemented a complete solution for a robot task set in a realistic home scenario. Within a limited amount of time, the students had to complete task understanding, data collection, model training, system debugging, and on-site demonstration. The competition placed high demands on algorithm design, visual perception, model deployment, hardware operation, and teamwork.
After intense competition, the IMCL team secured third place in the open developer division. Competing against teams from high-level research institutions and internationally recognized universities, the students delivered a clear technical plan, demonstrated solid professional foundations, and maintained stable on-site execution throughout the event.
2. Background of the Track and Team Performance
Physical AI is one of the most important development directions in today’s AI landscape. Unlike systems that operate only in virtual environments or on static data, Physical AI emphasizes interaction between intelligent agents and the real physical world. It requires a system to perceive its environment, understand tasks, plan actions, and execute them reliably on real devices.
For this reason, the track focused on real robot tasks rather than isolated benchmark performance. Teams had to rapidly build and refine technical solutions on site, balancing perception, model training, hardware control, and task execution under time pressure. This made the challenge a test not only of algorithmic strength, but also of engineering robustness and practical problem-solving ability.
The IMCL team applied its previous research experience in computer vision, deep learning, and intelligent systems to the competition scenario. Throughout the event, the team continuously refined its solution around task requirements and optimized system performance for the live demonstration. Their final result showcased strong hands-on engineering competence and calm execution in a demanding competitive setting.
It is also worth noting that the competition in the open developer division was extremely strong. The first- and second-place teams came from the Guangdong Institute of Intelligent Science and Technology and the University of California, Berkeley, respectively. Against this level of competition, the performance of the MPU team demonstrated both the strength and potential of Macao university students in global technology innovation competitions.
3. Message from the Laboratory
This award is an important reflection of IMCL’s sustained efforts in artificial intelligence, computer vision, intelligent perception, and Physical AI applications. The team’s research and practice have also benefited from support provided by Macao science and technology development projects, including grants 0014/2025/RIC, 0009/2025/ITP1, and 0004/2024/EIB1, which have created a strong foundation for student work in AI model training, visual perception algorithms, and intelligent system applications.
The laboratory has always encouraged students to pursue research training and engineering practice around real-world problems, and to take part in high-level technology competitions and cross-disciplinary innovation activities beyond the classroom. As a Macao-based platform with global reach, BEYOND Expo has also provided valuable opportunities for Macao university students to learn from outstanding teams from different regions, universities, and research institutions.
Looking ahead, IMCL will continue advancing research in artificial intelligence, intelligent medical computing, computer vision, and Physical AI, while encouraging students to identify and solve problems in real scenarios. Through this approach, the laboratory aims to keep strengthening innovation capacity and engineering practice, and to contribute to the development of science and technology innovation in Macao and the Guangdong-Hong Kong-Macao Greater Bay Area.
Once again, congratulations to Jiefeng Xie, Shengxi Chen, Jianchang Ruan, and Yue Sun on this excellent result.