Team KameRider Airost has won the Best Performance Award in the RoboCup 2023 Bordeaux Simulation Education Challenge. Congratulations to them! After their win in the Japan Open 2023, KameRider Airost’s second-generation team has reached greater heights in the international RoboCup 2023 in Bordeaux, France, held from July 5th to 9th. The team consisted of Anas Ismail Hassan Amer Aburaya, leader and 4th-year Bachelor of Electrical Engineering (Mechatronics); Mahmoud Abdelwase Mahmoud, 4th-year Bachelor of Electrical Engineering (Electronic); and Marco Bono, 4th-year Bachelor of Electrical Engineering (Mechatronics). The mentors were Loke Rui Kee, a 2nd-year Bachelor of Computer Science (Bioinformatics), and Mohammed Saeed Mahmood Ahmed, a 3rd-year Bachelor of Electrical Engineering (Mechatronics). The advisor was Dr Jeffrey Tan Too Chuan from MyEdu AI Robotics Research Centre.
The RoboCup Simulation competition evaluates the ability of intelligent robots to engage in natural and friendly communication with users while performing various support tasks in daily-life environments. The competition uses the SIGVerse simulator, which allows robots to make embodied and social interactions in virtual reality environments.
This time in Bordeaux, the @Home Simulation event consisted of 2 parts: workshop and challenge. The workshop is led by Rui Kee of @Home Education Simulation to train novice teams and prepare them for the challenge.
All activities occur virtually during the event, but the physical equipment is installed at the RoboCup venue – Bordeaux Exhibition Center in France. The online workshop and challenge are held concurrently with other on-site RoboCup leagues and challenges in the exhibition hall throughout the week.
The challenge consists of two parts: Handyman Task and Open Challenge. In a Handyman Task, the robot needs to understand human instructions to navigate to a specific room, pick up a particular object, and place it in a designated location. Two different trials are conducted, each with a different set of environments and instructions. The winner is the team with the highest score after completing both trials.
As for the Open Challenge, the team is required to conduct a 15-minute presentation that includes a robot simulation demo, followed by a 5-minute Q&A session. The main objective of the Open Challenge is to demonstrate the capabilities of each team’s robot system as a life support robot without any restrictions imposed by the task rules. Teams are encouraged to showcase their latest research and development on new approaches and applications for lifestyle support robots.
Team KameRider Airost performed exceptionally well in the challenge, achieving a remarkable score of 70 in the first trial and 100 in the second. During the Open Challenge, they presented their findings on computer vision, which they had deduced while working on the challenge. In addition, the team contributed to the workshop content by developing the Human Navigation Task. They used natural language processing, position determination, and distance calculation methods to generate simple instructions for humans to interact in a virtual environment and perform specific tasks.
We applaud their magnificent work and efforts. Despite all the difficulties, their dedication throughout the journey is truly remarkable. They overcame academic stress, learned, and developed on multiple platforms within eight weeks. This also proves that every effort counts. With the mentors from Japan Open, who were the first generation KameRider Airost members, and their code as a foundation, the second generation contributed to the development faster and made the overall architecture more mature.