Versatile Mechanical technology.

Uploaded on:
Category: Business / Finance
Above and beyond:
Slide 1

Mobile Robotics Julie Letchner Angeline Toh Mark Rosetta

Slide 3

Fundamental Idea: Robot Pose 2D world (floor arrangement) 3 DOF Very straightforward model—the trouble is in self-rule

Slide 4

Major Issues with Autonomy Sensor Inaccuracy Movement Inaccuracy Environmental Uncertainty

Slide 5

Problem One: Localization World guide Robot\'s underlying stance Sensor updates Given: Find: Robot\'s stance as it moves

Slide 6

How would we Solve Localization? Speak to convictions as a likelihood thickness Markov presumption Pose dissemination at time t adapted on: stance dist. at time t-1 movement at time t-1 sensor readings at time t Discretize the thickness by examining

Slide 7

Localization Foundation At each time step t: UPDATE every specimen\'s new area in view of development RESAMPLE the posture conveyance in light of sensor readings

Slide 8

Algorithms Markov confinement (least complex) Kalman channels (generally most mainstream) Monte Carlo limitation/molecule channels Same: Sampled likelihood dissemination Basic redesign resample circle Different: Sampling systems Movement presumptions

Slide 9

Localization\'s Sidekick: Globalization Localization without learning of begin area Credit to Dieter Fox for this demo above and beyond: "seized robot issue"

Slide 10

Problem Two: Mapping Robot Sensors Given: Find: Map of the earth (and verifiably, the robot\'s area as it moves)

Slide 11

Simultaneous Localization And Mapping (SLAM) If we have a guide: We can restrict! On the off chance that we can restrict: We can make a guide!

Slide 12

Circular Error Problem If we have a guide: We can confine! NOT THAT SIMPLE! On the off chance that we can restrict: We can make a guide!

Slide 13

Incorporate area/map vulnerabilities into a solitary model Optimize robot\'s exploratory way Use geometry (particularly inside) How would we Solve SLAM? Credit to Sebastian Thrun for this demo Major obstacle: relationship issue

Slide 14

Good diagram papers by Sebastian Thrun: "Probabilistic Algorithms in Robotics", 2000 "Robotic Mapping: A Survey", 2002 For the Interested Stanford course: cs225B Build a Markov Localization motor Run it on Amigobots to play soccer

Slide 15

Mobile robot sample: Underwater robots Localization is just helpful in case we\'re portable… … so how do these robots move? Up Next… Emergent Behaviors Mobile robots all the more intense in gatherings … … yet limitation is costly… … so what would we be able to manage without confinement?

View more...