Gen Z Landscape Architecture Graduate Becomes Robot Data Collector on CCTV
On September 13, Zhao Yingwei's phone buzzed nonstop. Former classmates and teachers he hadn't heard from in years suddenly messaged him. A netizen was "looking for Comrade Zhao," asking, "Is this really you?" The Zhao Yingwei interviewed on CCTV's Xinwen Lianbo was indeed him. He had no idea he would become an internet curiosity.
In August, he had just joined an embodied AI company when reporters happened to visit to film data collectors' daily work. Asked about his academic background, he answered truthfully: he studied landscape architecture. A netizen screenshotted the moment and posted: "I'd like to ask, how can someone with a landscape architecture degree join the robotics industry?" The post sparked discussion about unconventional paths into robotics.
Zhao is 25. After middle school, he entered a connected program between a vocational college and a Beijing municipal university—two years of high school, three of junior college, two of undergraduate. At high school graduation, he vaguely felt landscape architecture suited him. Junior college was like field trips: dendrology, floriculture, entomology, sketching, and simple software. In undergraduate, the engineering side emerged. The field has a "quick design" exam requiring landscape planning and design for a specific site within limited time. Though tiring, submitting the drawings gave Zhao a strong sense of accomplishment.
In his final undergraduate year, he learned many seniors had not stayed in design, and people were pessimistic about the field's prospects. Salaries were low, with long hours of drawing ahead. In 2024, he quickly signed with a construction company and after graduation became a manager stationed at real estate projects. The job avoided late-night drawing, offered room and board, and a decent salary. But colleagues were mostly around 40, and he lost touch with friends. Once on a video call, a friend joked: "You have so much 'workplace smell'—you look like you're 30 or 40."
Zhao joined part-time job information groups and sent resumes to several robotics companies for data collection positions. He learned such roles usually don't require a specific major, relying more on trial performance. During the trial, he had to start up a robot, connect VR, and use a controller with a gripper to pick up fruit. His first impression was novelty. He had always worked with concrete and steel; suddenly he was touching the frontier. In August, after passing the trial, he joined his current embodied AI company.
As a robot data collector, Zhao records data in many scenarios. Sometimes he wears a headband with a camera and holds a gripper, recording daily chores like tidying a restaurant or washing dishes frame by frame. These videos are important training data for robots. The job doesn't directly involve frontier algorithms. Zhao's idea is to "get on the bus first, pay later." He started with a part-time role, doing basic work, believing that if he proves his ability, he can later transfer to become a collection project manager. Every day after work, he self-studies Linux and basic Python, preparing for a technical role. He says there is no knowledge barrier in this era—online resources are vast, and AI can be a teacher. If you want to do it, nothing seems to be able to stop you.