Learning to Design and Construct Bridge without Blueprint

Autonomous assembly has been a desired functionality of many intelligent robot systems. We study a new challenging assembly task, designing and constructing a bridge without a blueprint. In this task, the robot needs to first design a feasible bridge architecture for arbitrarily wide cliffs and then manipulate the blocks reliably to construct a stable bridge according to the proposed design. In this paper, we propose a bi-level approach to tackle this task. At the high level, the system learns a bridge blueprint policy in a physical simulator using deep reinforcement learning and curriculum learning. A policy is represented as an attention-based neural network with object-centric input, which enables generalization to different number of blocks and cliff widths. For low-level control, we implement a motion-planning-based policy for real-robot motion control, which can be directly combined with a trained blueprint policy for real-world bridge construction without tuning. In our field study, our bi-level robot system demonstrates the capability of manipulating blocks to construct a diverse set of bridges with different architectures.

Authors

• Yunfei Li 0005
• Tao Kong
• Lei Li 0005
• Yifeng Li
• Yi Wu 0013

Keywords

• Bridges
• Shape
• Neural networks
• Reinforcement learning
• Reliability engineering
• Manipulators
• Task analysis
• Neural Network
• Deep Learning
• Robotic System
• Deep Reinforcement Learning
• Curriculum Learning
• Physical Simulation
• Training Policy
• Bridge Construction
• Attention-based Neural Network
• Building Blocks
• Flat Surface
• Representation Learning
• Path Planning
• Efficient Learning
• Reward Function
• Markov Decision Process
• Linear Layer
• Policy Learning
• Reinforcement Learning Algorithm
• Policy Network
• Construction Task
• Real Robot
• Bridge Design
• Average Success Rate
• Beginning Of Episode
• Planning Module
• Reinforcement Learning Agent
• High-level Planner
• Attention Block
• Target State