Robotics week5 Flashcards
Localisation / Maps-and-Path-planning (19 cards)
What is the difference between guidance and navigation?
Guidance: Directing the motion or position of something(robot)
Navigation: The process of accurately ascertaining one’s position and planning/following a route.
What is odometry(Dead reckoning) in mobile robotics
Odometry uses on-board motion sensors (e.g., wheel encoders, IMU) to estimate the robot’s change in position (or pose) over time by integrating velocity and orientation measurements, starting from a known reference location
Name at least four external positioning methods for indoor and outdoor localisation.
GPS, Wifi/Bluetooth beacons, iGPS, Ultra-wideband
What are the main limitations of relying on external positioning systems?
Accuracy constraints
Infrastructure dependency: Requires pre-installed beacons,satellites, or lasers -> unsuitable for unexplored or disaster areas
Security concerns: Signals can be spoofed or jammed
Exteroceptive only: Robots must also localise when external cues are unavailable.
Write the path-integration equations used in dead-reckoning equation
Given a distance d travelled at orientation theta: tri x = dcos(theta), tri y = dsin(theta)
then update the robot’s position at time t: xt = xt-1 + tri x, yt = yt-1 + tri y
why do odometry errors accumulate, what is the typical consequence?
Source of error
Accumulation: small errors in each step build up over time
Consequence: Large discrepancies from the true path - ranging from milimeters to kilometers
In the provided odometry example, if the robot starts at (0,0) and moves 3.1m at theta = -90, what is the odometry position and error if the true waypoint is at(0, -3)?
tri x = 3.1cos(-90) = 0, tri y = 3.1sin(-90) = -3.1
Estimated postion: (x1, y1) = (0+0, 0-3.1) = (0,-3.1)
Error:
e = |0 - 0| + | - 3.1 - (-3) = 0 + 0.1 = 0.1
What is optical flow, and how does it relate to localisation?
Optical flow is the apparent motion of pixel patterns between image frames caused by relative movement.
Relation to localisation:
Optical flow can estimate robot motion (translation and rotation) visually, contributing to visual odometry even without wheel sensors
What’s the image jacobian equation role in visual odometry?
Solve for camera velocities using visual input, critical for visual odometry
What are the two main types of robot map representations?
Topographic (metric): Quantitative, uses absolute positions
Topological: Qualitative, uses nodes and edges to represent connections between places
What are typical sensors used to build topographic maps?
LIDAR
Vision(camera)
Infrared
Ultrasonic
What is an occupancy grid
A discrete raster format map where each grid cell indicates unexplored, free space, obstacle.
What is the main difference between topographic and topological maps?
Topographic: Detailed, metric, need accurate odometry
Topological : Abstract, symbolic, more scalable but depends on place recognition.
what is hybrid map?
A map that combines topographic, topological, and semantic levels to balance accuary, abstraction, and human-meaningful data
What is cost-map in robotics
A hybrid map where each cell hold a cost value based on multiple factors with obstacles, distance, or dancer - used for path planning
What does Dijkstra’s algorithm optimise?
It finds the minimum cost path from the start node to all other nodes by optimising: f(n) = cost from start to node n
What is the drawback of Dijkstra’s algorithm in large graph??
It performs flood search and is computationally expensive, exploring many irrelevant paths.
What is the A* search formula and its components?
f(n) = g(n) + h(n)
g(n) : cost from start to node
h(n): heuristic estimate from node to goal(e.g, straight-line distance)
