Interdisciplinary engineering research and education covering fundamental engineering principles, systems thinking, engineering design methodology, and cross-cutting engineering challenges
3 papersThis paper introduces MEVITA, an open-source bipedal robot built entirely from e-commerce components assembled using sheet metal welding. Reinforcement learning in simulation, combined with Sim-to-Real transfer, allowed MEVITA to demonstrate robust walking behavior across varied terrains. The design emphasizes minimizing the number of components and ensuring easy assembly.
This study developed a hybrid AI model to predict how much the ground sinks during subway tunnel construction. Testing it on a real project in Guangzhou, China, they found it was pretty accurate, especially for small settlements, and also figured out which construction factors matter most (like the pressure at the front of the tunnel boring machine).
This study develops an improved PSO-BP neural network model to predict surface settlement during rectangular pipe jacking tunnel construction. The model, tested on a real-world case study, outperforms traditional methods by accounting for adaptive mutation and incorporating inertia weight and mutation particles for enhanced accuracy and stability. The findings suggest this model can be valuable for similar projects.