SWARM ROBOTICS
Laurynas Veržukauskas
REQUIREMENTS
EACH unit
-
Autonomous
-
Scalable
-
Homogeneous
-
Collaborative
- Localized
PHEROMONES
Gradient
Algorithm
- Brownian motion until food or pheromone trail is located
- If pheromone trail is located - follow it
- If nest reached - follow it to opposite direction
- Pick up food and take it back to the nest leaving pheromone trail
- Go back to food using available trail, leaving pheromone trail
VIRTUAL PHEROMONES
-
Get info from nearby robots
-
Calculate gradient map
-
???
-
Profit
BASIC BEHAVIORS
- Dispersion
-
Aggregation
- Pattern formation
- Collective movement
- Task allocation
- Source search
- Collective transportation of objects
- Collective mapping
Emergence concept
ARCHITECTURE
SIMPLE*
-
Locomotion
-
Communication
- Extras
* - directional transmitters and receivers, limiting signal strength, limited objects recognition, etc.
LOCOMOTION
Rotors
Wheels
Flagella
Piezoelectric
...
COMMUNICATION
GLOBAL
Not required and unwanted for base case
LOCAL
Critically important (only way to comm. with swarm)
Directional (provides sense of direction)
COMMUNICATION
IR vs RF
INTERFERENCE
Noise amount
Optimized Not optimized
OPEN ISSUES
Hardware limitations
Lack of design methodology
Lack of testing methodology
Lack of formal verification/guarantee of properties
Security issues (vast separate topic)
...
ADVANTAGES
Scalable
Simple architecture
Robust swarm
Cheap
APPLICATIONS
Commercial
not yet
Theoretical
Any search tasks (search & rescue, exploration)
Surveillance (building security, battlefield)
As platform for something (distributed antennas)
Swarm robotics
By Laurynas Veržukauskas
Swarm robotics
Slides for short presentation about swarms robotics
