Newshot

Average age 25: Gen Z-led team scores highest at first Ant Lingbo Embodied Large Model Challenge

AI summary

On November 16, the finals results of the first Ant Lingbo Embodied Large Model Challenge were announced. The "TARS" team, made up of five contestants from the Futian Laboratory, Northwestern Polytechnical University, Pudu Robotics, and Tianjin University, won first prize; the "EVA" team, made up of three contestants from Huazhong University of Science and Technology, the Tsinghua Shenzhen International Graduate School, and Hubei Yuanyi Intelligent, and the "SST" team, made up of three contestants from Tianjin University, Peking University, and the Institute of Automation of the Chinese Academy of Sciences, won second prize; the "Vida," "FDU_SONIC," and "ai-robot" teams won third prize; another 9 teams received excellence awards, and 3 teams received special observer group awards. The TARS team won first prize at the first Ant Lingbo Embodied Large Model Challenge. Image source: interviewee. After 48 hours of real-machine competition, the robots had to pass three stages. On November 14, 15 finalist teams gathered at the Shanghai Zhangjiang Science Hall. Contestants came from Beijing, Shanghai, Hangzhou, Shenzhen, Chengdu, Wuhan, and other places, and some contestants returned speci

Translation(model translation, not the source's own text)

Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge

On November 16, the results of the finals of the first Ant Lingbo Embodied Large Model Challenge were announced.

The "TARS" team, formed by five contestants from the Futian Laboratory, Northwestern Polytechnical University, Pudu Robotics, and Tianjin University, won the first prize; the "EVA" team, formed by three contestants from Huazhong University of Science and Technology, the Shenzhen Research Institute of Tsinghua University, and Hubei Yuanyi Intelligent, and the "SST" team, formed by three contestants from Tianjin University, Peking University, and the Institute of Automation of the Chinese Academy of Sciences, won the second prize; the "Vida", "FDU_SONIC", and "ai-robot" teams won the third prize; another 9 teams won the Excellence Award, and 3 teams won the Observer Panel Special Award.

Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge
Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge

The TARS team wins first prize at the first Ant Lingbo Embodied Large Model Challenge. Image source: the interviewee

48 hours of real-machine competition, and the robots must pass three tests

On November 14, 15 finalist teams gathered at the Shanghai Zhangjiang Science Hall. The contestants came from Beijing, Shanghai, Hangzhou, Shenzhen, Chengdu, Wuhan, and other places, and some contestants even returned from Vancouver specifically to compete. At 9 a.m., as soon as the challenge tasks were announced, the teams dove into their workrooms and began a 48-hour real-machine hackathon.

In the competition, each team used the same open-source pre-trained model, LingBot-VLA 2.0, as the technical foundation, and carried out real-machine challenges using unified robotic hardware, inference compute, and task environments. There were three challenge tasks: first, using a single arm to grasp fruits and vegetables; second, using a single arm to grasp cosmetics and place them into a transparent box; and third, using both arms together to sweep up desktop trash and then pour it into a storage basket.

Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge
Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge

At the competition site, a contestant operates the robotic arm to complete the dual-arm collaborative trash-sweeping task. Image source: the interviewee

These actions are very common in daily life, but once they are performed by a robot, every step must withstand scrutiny. From recognizing objects, grasping firmly, and placing accurately—especially interaction with transparent objects—to the two robotic arms coordinating with each other, the judges comprehensively assess the quality of action completion and the task success rate. The finals consisted of three rounds of real-machine evaluation, and each team's highest total score from one round was counted toward the ranking.

In the end, the "TARS" team won first place with 82.5 points. This is a young team with an average age of only 25, and the team name comes from the loyal and humorous robot in the film Interstellar. The team leader, 吴昊楠, is a post-00s student about to graduate with a master's degree from Tianjin University. During the roadshow segment, he presented the team's technical thinking and strategy: by simplifying complex multi-object scenes into single-object grasping, they effectively avoided the perception bottleneck of VLA models in cluttered environments; they chose small-scale, high-precision data recollection and a two-person collaborative collection mode to avoid the systematic errors of robotic arms in the physical world and to prevent human operators' physiological comfort from affecting the naturalness and effectiveness of robot motion trajectories.

李永露, associate professor at the School of Artificial Intelligence of Shanghai Jiao Tong University, believes that "what the Embodied Large Model Challenge has precipitated through competition is not only honor, but also a set of engineering experience and methodology that has withstood real-world testing." By unifying the foundation model, hardware platform, and practical tasks, the competition has built a "controlled-variable experimental field" for university scholars, industry engineers, and independent geeks. Every boundary condition of a failed grasp, every set of precisely calibrated supplementary data, and every real-world feedback after model tuning can reduce the trial-and-error cost for those who follow and push the robot "general brain" into the next real physical setting.

Adhering to embodied nativity, Ant Lingbo releases a developer ecosystem plan

On the same day, Ant Lingbo also held its first Developer Conference, where globally renowned scholars and frontline industry practitioners shared the latest research and practice in embodied large models and discussed topics such as the difficulties in real-machine deployment and how to jointly build the developer ecosystem.

In the scientist technical lecture segment, Shen Yujun, Chief Scientist of Ant Lingbo, elaborated on Ant Lingbo's embodied-native technical philosophy, which starts from real needs. He said that a smart robot brain faces three challenges: inputs contain errors, outputs must withstand the test of physical interaction, and tasks in the physical world will not wait for the model; being able to solve a robot's problems does not mean being suitable for robots. In response, Ant Lingbo designs model capabilities around physical operations and builds an embodied-native brain, enabling robots to think more clearly, feel more sensitively, understand more clearly, and act more efficiently.

Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge
Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge

Shen Yujun, Chief Scientist of Ant Lingbo, elaborates on Ant Lingbo's embodied-native brain. Image source: the interviewee

At the conference, Ant Lingbo announced its developer ecosystem plan, which aims to invite 20 of Ant Lingbo's most valuable experts to support teaching and academic innovation and jointly cultivate embodied intelligence talent; at the same time, it will recruit 40 Ant Lingbo developer experts to support technical practice and professional growth and jointly build the developer ecosystem.

Embodied intelligence is still an industry in the ascendant. Zhu Xing, CEO of Ant Lingbo Technology, emphasized in an interview with Leiphone.com (official account: Leiphone.com) and other media that the industry is still iterating rapidly and has not yet reached the Scaling Law moment for embodied intelligence. However, with the validation of some technologies in deployment and collaboration across industry, academia, and research ecosystems, the evolution of embodied intelligence will also accelerate.

This Ant Lingbo Embodied Large Model Challenge consisted of two stages: an online algorithm preliminary round and an offline 48-hour real-machine hackathon final. It attracted 1,855 teams and 3,252 contestants from 13 countries around the world, covering 648 universities, research institutes, technology companies, and embodied intelligence teams, making it the largest embodied large model challenge in the industry.

  • Fable 5.2 "collides" with Ant Lingbo's tactile model, CEO Zhu Xing: the two ...
  • The industry's largest embodied model challenge reaches the finals: 15 teams will showcase ...
  • Chinese open source moves into the physical world, Ant Lingbo brings its robot general brain to ...
  • A record more than 1,800 teams enter, Ant Lingbo's embodied model challenge is about to ...
  • DEEP Robotics releases a new quadruped robot, achieving a 50Kg payload and operation in complex terrain
  • Average age 25: a post-00s young man leads his team to the highest score at the first Ant Lingbo Embodied Large Model Challenge
  • When charging is as fast as refueling, the boundaries of renting a car for road trips are expanded
  • Qijing GT7 Ultra Comfort Edition pure electric/extended range officially launched! Down payment from 66,600 yuan, get an L3-level architecture shooting brake
  • Doubao Work exclusively launches the "Workbench" feature, initially covering 21 professional scenarios including finance, healthcare, and education
  • Korean consumers fall in love with AliExpress Double 11! Chinese brands' transaction volume surges 4 times
  • Korean consumers fall in love with Double 11! Chinese brands' transaction volume surges 4 times on AliExpress
  • Freedom of model choice! Qianwen Office launches 10 models including Qwen4, Kimi-K3.1, and GLM-5.3

The translation is machine-generated and individual terms may be off. Nothing is rewritten or cut; the original is authoritative.

Published by 雷锋网
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗