macro_dice.md

Digital Material Modeling and Simulation

Case study of the design and optimization of Macro-DICE.

Hardware Implementation:

• VoxRob
• rover

---

Steps

1. Find Goal
   • CNN to learn target (instructions) from Image
   • Progress
   • Create/get data to train
2. Auto Encoder to find latent representation for it
3. Control Strategy based on the
   • Train using the simulator + WANN
   • Progress
4. Fit Arbitrary Graph/data flow on DICE
   • Progress
5. Assembly and Testing
   • Meso-dice assembly
   • Robotic assembly

---

Simulation Environment

Rover

• Check Control Function Graph

• Simulation Demo
• Progress

Wal

• Progress

---

Optimization

• Two level optimization
  • Shape (body) => Hierarchal Local Online Shape Optimization
  • Control (mind) => AI that grows
  • Simultaneous optimization of shape and control

---

Robotics Design

Using the MetaVoxel simulation tool:

---

Assembly

• Option for different kinds of assembly (swarm assembly)