discrete-assembly-sim

Discrete Assembly Simulation: Advanced Coordination for Lattice-Building Robot Swarms

This repository offers an innovative coordination stack designed for robot swarms engaged in lattice construction. It provides seamless functionality from blueprint creation to validated build orders, while facilitating multi-robot choreography and ensuring robust error correction. A crucial feature of this project is its ability to maintain a comprehensive replayable log of each operational run.

Small robots, derived from the MIT BILL-E and NASA ARMADAS lineage, effectively navigate the voxel structures they are constructing. They autonomously retrieve components from a storage depot, execute tasks based on a solver-validated build sequence, manage spatial negotiations using a shared reservation system, and conduct thorough inspections. If defects are identified, they remove and replace faulty components without interrupting the assembly process in the rest of the swarm.

Key Features

• Open Coordination Stack: Designed for flexibility, the system allows users to customize robot capabilities according to their specific configurations.
• Error Correction Mechanism: The project employs a methodology similar to digital data error correction, ensuring high reliability in assembly despite individual component failures.
• Three-Dimensional Coordination: Unlike 2D environments, the 3D structure provides real parallelism, allowing robots to navigate without collisions, thereby enhancing throughput.
• Compatibility with Various Robots: The design principle asserts that the coordination logic remains invariant regardless of the hardware utilized by the robots, simplifying integration efforts.

Performance Insights

Research shows a significant improvement in assembly efficiency in three-dimensional constructs compared to two-dimensional operations. For instance, while a two-robot deployment in 2D yields decreased performance due to choke points, deploying multiple robots in 3D facilitates better spatial utilization and task completion rates.

Error Correction Yield

The simulation also demonstrates that a builder engaged in inspection and repair can maintain near-perfect yield even under adverse conditions, which significantly outperforms conventional methods where inspection is not performed.

Usage Example

To implement this simulation in your project, simply install via the following command:

pip install git+https://github.com/jdyar/discrete-assembly-sim.git  

Users can initiate demo simulations to visualize how different robots interact within various lattice configurations. For example:

python examples/my_robot.py  

This command runs a simulation showcasing a personalized robotic assembly strategy.

Future Directions

Future development aims to address dynamic robot coupling, expanding the range of configurations and enhancing the scalability of the coordination stack. Moreover, ongoing efforts will include incorporating typed part families and additional structural checks.

This project seeks collaboration and invites contributions to improve the functionality and extend the application range of the discrete assembly simulation framework. Contributions such as bug reports or alternative robot integrations are highly appreciated.