The closing ceremony of the “2024 X-Challenge Disruptive Innovation Camp” (X-Camp), organized by the X-Institute, Shenzhen, and jointly hosted by the Center of Double Helix (CDH) of the Tsinghua Shenzhen International Graduate School (SIGS) was successfully held recently at the SIGS.
During the closing ceremony, Jun Liu, rector of City University of Macau, and Quanshui Zheng, founding dean of X-Institute, signed a cooperation agreement to jointly establish a platform for cultivating top-notch innovative talents. Two students from the Faculty of Data Science, Chen Tianyu and Zhang Yanze, participated in this year’s X-Camp. Among them, Chen Tianyu and his team received the Disruptive Innovation Project Award for their research project titled “A Hybrid Self-Healing Drone Swarm Forest Firefighting System Based on Intelligence Technologies.”
The signing of the agreement between CityU and the X-Institute signified the beginning of comprehensive strategic cooperation to implement a top-tier innovative talent cultivation model, paving the way for high-quality educational development and an ecological environment for nurturing talent.
Rector Jun Liu, along with experts and scholars including Guo Dongming, academician of the Chinese Academy of Engineering, Chen Yeguang, academician of the Chinese Academy of Sciences, Chen Xiaodong, elected fellow of Royal Society of New Zealand, Yang Jun, fellow of the Canadian Academy of Engineering, and Academician Quanshui Zheng, engaged in an in-depth exchange and explored models for cultivating top-tier talent and future collaboration plans.
As an innovative and research-oriented learning camp, this year’s edition reached out to students worldwide. Approximately 140 high school students and 220 college students were selected for a 20-day research and innovation experience at the X-Institute.
The participating students presented nine innovative research group projects, all based on real and cutting-edge challenges from around the world. Academic guests, including leaders and experts from major universities, innovation education institutions, and various professional fields, attended the project presentation sessions and engaged in in-depth discussions with students during the Q&A session and the Gallery Walk project exhibition.
CityU student Chen Tianyu and his team presented their innovative project research findings titled “A Hybrid Self-Healing Drone Swarm Forest Firefighting System Based on Intelligence Technologies.” This project utilized hybrid sensors and fuel-electric drones with hierarchical model for drone control, scheduling and an election algorithm for selecting the master drone. Under the aerial view of the master drone, the team conducted simulations for obstacle avoidance.
Compared to traditional distributed or centralized systems, this solution addressed issues such as complex challenges or the loss of the master drone. For single-drone obstacle avoidance, it employed deep reinforcement learning (DRL) with a diffusion model as the basis for its strategies. Better performance was demonstrated in complex environments compared to conventional algorithms.
Finally, the “Lifelong Federated Reinforcement Learning Framework” was proposed for training intelligent systems. This framework leveraged high perception achieved through hybrid sensors, self-decision-making from DRL, distributed training, and real-time computation through edge computing to enhance the training of the intelligent systems. It also utilized the digital twins for data collection to facilitate subsequent simulation training.
[foogallery id="66249"]
Source: Faculty of Data Science, X-Institute, Shenzhen
