Collective Hunting using E-Pucks BIEHL Martin HADBI Nadir - - PowerPoint PPT Presentation

▶
collective hunting using e pucks
SMART_READER_LITE
LIVE PREVIEW

Collective Hunting using E-Pucks BIEHL Martin HADBI Nadir - - PowerPoint PPT Presentation

Jul 02, 2023 413 likes •731 views

Collective Hunting using E-Pucks BIEHL Martin HADBI Nadir Assistant : RAEMY Xavier Goal of the project Simulation of two predators and a twice as fast prey in a bounded arena Only solution to catch the prey ? Collaborate

slide-1
SLIDE 1

Collective Hunting using E-Pucks

BIEHL Martin HADBI Nadir Assistant : RAEMY Xavier

slide-2
SLIDE 2

Goal of the project

Simulation of two « predators » and a

twice as fast prey in a bounded arena…

Only solution to catch the prey ?

Collaborate !

We explored two ways of collaboration :

Observing the behavior of other predators and

prey -> « Image processing » (like wolves for ex.)

Information exchange -> Wireless

communication

slide-3
SLIDE 3

Image Processing

Simulated in « Webots » To facilitate image processing:

« Perfect World » : Predators are green, Prey is red, Arena is blue

slide-4
SLIDE 4

Strategy

Some examples of situations…

slide-5
SLIDE 5

Searching (no com)

slide-6
SLIDE 6

Following (no com)

slide-7
SLIDE 7

Following (no com)

slide-8
SLIDE 8

Following (no com)

Sees

  • ther

predator in front, and follows him

slide-9
SLIDE 9

Waits for his leader to take a turn i.e. ∆x surpasses a minimum value

Following (no com)

∆x distance travelled between two camera shots

slide-10
SLIDE 10

Leader turns when prey changes direction Starts spinning in place in direction

  • f his

leaders turn.

Turning (no com)

slide-11
SLIDE 11

Starts spinning in place in direction of its leaders turn, until prey and predator are

  • ut of sight…

…and a little longer to get good results. (CRUCIAL POINT)

Turning (no com)

slide-12
SLIDE 12

Then it starts straight into the final direction.

Trapping (no com)

slide-13
SLIDE 13

Trapping (no com)

slide-14
SLIDE 14

Maybe its a successful trap …

Trapping (no com)

slide-15
SLIDE 15

…maybe not.

Trapping (no com)

In this case the trapping predator sees the prey pass by in front of him and thus he starts following it.

slide-16
SLIDE 16

…maybe not.

Relay of following task (no com)

The formerly following robot now sees the other one in front of him and follows him.

slide-17
SLIDE 17

Trapping (no com)

And turns …traps … follows …

slide-18
SLIDE 18

Problem (no com)

Will spin in wrong direction, because he depends on his partners movement.

slide-19
SLIDE 19

State Machine (image processing)

slide-20
SLIDE 20

Wireless Communication

Emitter/ Receiver communication system

with ideal range

Relay No longer dependent on the movement of

the other predators

slide-21
SLIDE 21

Searching (with com)

slide-22
SLIDE 22

Following (with com)

States are constantly broadcasted. One predator will see the prey first. And follow it.

slide-23
SLIDE 23

Trapping (with com)

States are constantly broadcasted. Knowing there already is a follower, the

  • ther predator

now follows the prey and decides to turn and trap depending on the preys behaviour. (∆x)

slide-24
SLIDE 24

Relay of following task (with com)

States are constantly broadcasted. Again we relay the following task to the formerly trapping predator.

slide-25
SLIDE 25

Following (with com)

States are constantly broadcasted.

slide-26
SLIDE 26

Following (with com)

States are constantly broadcasted.

slide-27
SLIDE 27

State Machine (communication)

slide-28
SLIDE 28

Results

  • How efficient is our algorithm ? Does the

collaboration work ? Scalable ?

  • > Compare it to others controllers :

a)

Random Predators

b)

2 followers

c)

2 anticipators

d)

Specialist predators (1 follower + 1 anticipator)

slide-29
SLIDE 29

Results

  • Predators with comm vs 2 anticipators
slide-30
SLIDE 30

Results

Comm Robots vs No-Comm Robots

slide-31
SLIDE 31

Conclusion

Scalability : ? Limitations :

Implementation to real robots -> Image

processing, communication -> Noise !

Further improvements

GA to find optimal parameters Behavior of the prey Wider arena