During my TCL internship, I worked on Ai Me’s motion tests, investigated servo jitter, tested a direct-drive brushless alternative and revised internal layout to reduce vibration coupling.
PROJECTAi Me Companion Robot
MY ROLEMotion testing, actuator prototyping and internal structure
CONTEXTTCL Jul–Sep 2025
Inspect full-size image ↗FIG. 01Public TCL product imagery; my R&D contribution is summarised at a high level under NDA.
01 / OVERVIEW
The project in a minute.
THE OUTCOME
Contributed motion-control and structural work to the demonstration prototype shown at IFA 2025.
Help a companion robot move with less vibration and more natural expression without exposing confidential product details.
My contribution
Motion testing, actuator prototyping and internal structure. My CV records motion-test routines, actuator diagnosis and prototyping, internal-layout changes and interface work with the textile team.
Read this in context
Placement work on motion and prototype integration for Ai Me. The overview uses public product imagery; internal drawings and measurements are not published.
Open a chapter to follow the work and the choices made along the way.
01 / ENGINEERING
Investigate the mechanism behind the jitter
+
Motion tests linked servo jerk to control-step interpolation and gearbox backlash. I iterated a direct-drive brushless alternative under the lightweight chassis constraints.
Placement work on motion and prototype integration for Ai Me. The overview uses public product imagery; internal drawings and measurements are not published.
Actuation
Motion tests and prototyping
Diagnosis and a direct-drive alternative under real actuator constraints.
Integration
Internal structure and interfaces
Layout iteration and collaboration with the textile team.
IFA 2025
Demo prototype contribution
Motion-control and structural contributions, as documented in my CV.
Source notes & scope 1
CV James Wu — Robotics / Chinese embodied robotics CVTCL internship section
Role, engineering contributions and IFA 2025 demonstration context.
↗FIG. 02Public product imagery. Confidential R&D artefacts are not reproduced here.
Hands-on experience connecting physical robots, perception, control and services through task demonstrations, teleoperation, navigation experiments and hardware work.
My partTechnical Support Engineer Intern · robot applications and testing
A standalone temporal model outperformed a current-force heuristic offline. The integrated ForceVLA risk head also learned the proxy risk signal; reliable action improvement and closed-loop performance remain unverified.
Independent projectIndependent research: trajectory ranking, simulation and risk-head training