Student opportunities

Evolutionary Robotics

Voxel-based soft robots (VSRs)
- Cooperative co-evolution of body and brain of VSRs
- Learning techniques for the controllers of VSRs: reinforcement learning
- Auto-assembly of VSRs (ongoing, but expandable)
- Grammar-based generative encoding for developing (phenotypically plastic) modular robots (ongoing)
- Resolution-agnostic representation for evolution of closed-loop controllers of VSRs
- Social/cultural/imitation learning in VSRs
  - Physical vs. observed imitation
- Distributed world models for reconfigurable soft robots
- Impact on evolvability of the controller of the replacement of rigid parts with soft ones in modular robots
- Hierarchical (module role-based or environmental context-based) policies for modular robots
Real robots
- PushGP on mobile robots (Thymio-II): a machine-friendly representation for robotic agents controllers
Hybrid (soft & rigid) robot simulation and optimization
- Design and development of an interactive software tool for definining robot morphologies

Optimization of the policies of a few ruling agents in a multi-agent systems
Evolutionary optimization of synapsis-wise and reward-driven autoadaptation rules
Reconfigurable Neural Cellular Automata
Autogenerating neural-networks

Two- or multi-stage quality-diversity evolution with a representation based on anchor solutions
Co-evolutionary map-elites
Map-elites with trajectory-based, instead of point-based, pupulation arrangement
Cellular automata-based evolutionary algorithms
Diversity promotion based on neutrality graph