Adaptive sampling in environmental Robotics Mohammad Rahimi, Gaurav sukhatme, William Kaiser, Mani Srivastava, Deborah Estrin
Motivation… NIMS: Networked Infomechanical Systems
Internet Telephone/ISP Data Base WRCCRF Crane Site Dry Shack NIMS Node Met Node (Ta, RH, PAR) Solar Cell Battery Pack Power Distribution Cable Visible Imager With Pan/Tilt Actuator 47 m 50 m
Science Objectives C H2OH2O Q C 13 /C 12 T, RH, Wind Growth
NIMS Prototype Deployment
Problem Creating a dynamic Map of the environment Based on the carrying sensors (attributes)
Approach The robot (shuttle) is an agent gather Geostatistics information Refresh those statistics as fast as possible
Digitizing robots world 0,0 Cell size that we call it a pixel is a x*x. pixel is the distance that shuttle moves atomically Obstacles Shuttle Patrol Area
Assumptions Shuttle is certain about the location Sensor reading error is zero Environment is static in circuit convergence time Warning to the user to reduce coverage or expected accuracy otherwise
Sampling Policy Stratified Sampling Divide the population into subpopulations Extremely better performance with some degree of apriority domain knowledge Random sampling Mean proportional to cell size
Feedback Using variance of data to classify a region Vaiance/Mean < Expected error or Variance < Sensor Noise
Divide and Conquer Stratify the current cell into four μ = α * cell size (μ is mean of step size) Collect data in current cells (Random) Calculate the variance Iterate until variance is below threshold
Closed Loop System Estimation error Stratification Policy + - Map Acceptable error Reading points
Result Of the Algorithm n Log (n) n Quad-tree Map of the variance of the environment Shuttle step-size is random but proportional to how deep in the tree it is
Initial Results
Wish List Adding time domain Static sensors as sample support