Robotics/2022 MRS Camp

Oflactory

Young_Metal 2022. 8. 4. 00:45

Silkwarm

 

Bee

 

Dog

 

Ant

 - Cooperative foraging 

 

 

Robots with noses why...?

 - Night - Lighting condition. 

 - Sense - chemically interactive. 

1. Territory marking and avoiding

2. Search and Rescue

  Urban disaster, Wilderness search

3. Chemical Plumes

Plume movement  ~~ Odur dispersion

sensitive, selective, fast response (Gas sensor)

 

Fast -> 

 

Sampling -> Sensor -> Data Analysis with feedback and adaptation

Chemosensor MOS, Ion mobility...

Anemometer

Wind vanes -> vane that aligns with the wind

 

Search problem

 - finding the odour sources

how to sample the environment in order to reach

3 sub problems

 - sampling the space....ㅎㅎ

 

Sample -> Plume detect-> Track ->Source detected -> Declare

 

Sampling the space

Sequential search

 Regular trajectories

Distance among trajectories considers the probability of detection

 

While uncertainty > threshold do

1. Evaluate set of actions

2. Move to best sampling point

3. Sense

4. Update

 

2007. Cognitive approaches infotaxis POMDP

 

meta-heuristic approaches for odour search

 

Q memory 줄이기 가능하냐? Interaction gradient source. Advantage in oflactory sensing in application

A : Realize infect in reality, gradient approach. Swarm do detect source... sensing all the space.. combine the agent gradient? Relatie....뭐라 카노

 

'Robotics > 2022 MRS Camp' 카테고리의 다른 글

MRS last one  (0) 2022.08.11
Resilient Autonomy as a goal for robotics  (0) 2022.08.04
Window에서 Ubuntu로 Gazebo랑 Rviz가 안뜰때, RViz 윈도우에 띄우기  (0) 2022.08.03
Multi Robot System  (0) 2022.08.03
DARPA SubT  (0) 2022.08.03