Siming Sun

Topic: Rapid Surface Reconstruction Based on Bayesian Meta-Learning for Real-Time Obstacle Prediction and Avoidance

Completed Task 1: Updating the existing neural network models from TensorFlow 1.x to PyTorch

Completed Task 2: Building a new goal-seeking and exploration framework based on SDF, MPC and CBF

Completed Task 3: Testing the real-time performance of new models in the simulation environment.

Result: The new model enable rapid adaptation to dynamically changing environment.

Note: 1.3

Topic: Language-conditioned Learning for Robotic Manipulation with Linguistic Symmetry

Completed Task 1: Collecting and surveying on LLM

Completed Task 2: Simplifying code implementation and reduce redundancy

Completed Task 3: Building a buffer to save symmetric language instructions and trajectories

Result: The average success rate increased by 10%, specific tasks such as 'push left' and 'reach right' increased from 35% to 60%.

Note: 1.0

Completed Task 1: Code implementation of initial value problem using Matlab

Completed Task 2: Code implementation of stable and unstable heat conduction using Matlab

Completed Task 3: Code implementation of Linear and nonlinear Equations

Topic: Simulation of a bottling plant using Matlab/Simulink

Completed Task 1: Modeling of the mechanical components of the system

Completed Task 2: Virtual recording of bottle data within the system

Completed Task 3: Development of a control logic for the bottling plant in Stateflow

Completed Task 1: Developed technical documentation for internal and external stakeholders, ensuring effective communication of complex concepts.

Completed Task 2: Enhanced product development by conducting extensive research and analysis on industry trends.

Completed Task 3: Implemented data-driven approaches to decision-making by utilizing statistical analysis tools during product development.