## Ego-motion estimation and collision detection for omnidirectional robots (2006)

Venue: | In RoboCup 2006: Robot Soccer World Cup X |

Citations: | 3 - 1 self |

### BibTeX

@INPROCEEDINGS{Lauer06ego-motionestimation,

author = {Martin Lauer},

title = {Ego-motion estimation and collision detection for omnidirectional robots},

booktitle = {In RoboCup 2006: Robot Soccer World Cup X},

year = {2006},

publisher = {Springer}

}

### OpenURL

### Abstract

Abstract. We propose an algorithm to estimate the ego-motion of an omnidirectional robot based on a sequence of position estimates. Thereto, we derive a motion model for omnidirectional robots and an estimation procedure to fit the model to observed positions. Additionally, we show how we can benefit from the velocity estimates deriving an algorithm that recognizes situations in which a robot is blocked by an obstacle. 1

### Citations

715 |
Applied optimal estimation
- Gelb
- 1974
(Show Context)
Citation Context ...ment only from the wheel encoders is not reliable. As well, optical flow analysis [4] fails since optical flow is misleading in an environment in which other objects move. Kalman filtering techniques =-=[5]-=- suffer from the assumption of a linear motion model. Even the first order approximation which is made in the extended Kalman filtering approach may be misleading if the robot’s angular velocity is la... |

76 |
Navigating Mobile Robots: Systems and
- Borenstein, Everett, et al.
- 1996
(Show Context)
Citation Context ...trajectories that can be driven by such a robot. Assuming a robot with differential drive first and assuming constant angular velocity ω and linear velocity v the robot drives on an arc. According to =-=[8]-=- the expected position (x,y) and robot heading φ at time t is: φ(t) = ωt mod 2π (1) x(t) = v (cos (ωt) − 1) ω y(t) = (2) v sin(ωt) ω (3) Here, x and y are referring to the robocentric coordinate syste... |

50 |
Estimating motion in image sequences
- Stiller, Konrad
- 1999
(Show Context)
Citation Context ... the Middle Size League.s2 Estimating the Ego-Motion 2.1 Principle Idea As mentioned before, estimating the robot movement only from the wheel encoders is not reliable. As well, optical flow analysis =-=[4]-=- fails since optical flow is misleading in an environment in which other objects move. Kalman filtering techniques [5] suffer from the assumption of a linear motion model. Even the first order approxi... |

16 | Calculating the perfect match: An efficient and accurate approach for robot self-localization
- Lauer, Lange, et al.
- 2005
(Show Context)
Citation Context ...ther than a line. In recent years, some reliable and accurate self-localization algorithms have been proposed which allow to estimate the robot’s pose with an expected error of only a few centimeters =-=[6, 7]-=-. These approaches already combine different sources of information like wheel encoders, camera images and laser range scans to eliminate sensor noise. Thus, using these aggregated information makes t... |

9 | G.: Laser-Based Localization with Sparse Landmarks
- Strack, Ferrein, et al.
- 2006
(Show Context)
Citation Context ...ther than a line. In recent years, some reliable and accurate self-localization algorithms have been proposed which allow to estimate the robot’s pose with an expected error of only a few centimeters =-=[6, 7]-=-. These approaches already combine different sources of information like wheel encoders, camera images and laser range scans to eliminate sensor noise. Thus, using these aggregated information makes t... |

5 |
Motion estimation of moving objects for autonomous mobile robots
- Lauer, Lange, et al.
- 2006
(Show Context)
Citation Context ... to use an adaptive strategy that reduces n when the robot accelerates or decelerates and increases n when it drives with constant velocity. This idea is very similar to the approach that was used in =-=[9]-=- to estimate the ball velocity. Figure 1 shows the velocity estimates for a run on a real robot of the Brainstormers Tribots RoboCup team. Obviously, when accelerating the robot, there is some delay u... |

2 |
M.: Predicting away robot control latency. In: RoboCup 2003: Robot Soccer World Cup VII
- Behnke, Egorova, et al.
- 2003
(Show Context)
Citation Context ...e crucial to be able to interact with the rolling ball and with teammates and opponent robots. E.g. taking into consideration the internal delays in sensors and motor controllers of about 100 − 200ms =-=[1]-=- and robots that drive with 2 − 3 m s the robot can cover a distance of at most 60cm meanwhile. Furthermore, assuming maximal deceleration of 4 m s 2 the stopping distance is longer than 1.1m. While m... |

1 |
A.J.: The analysis of rigid body motion from measured data
- Shiflett, Laub
- 1995
(Show Context)
Citation Context ...lgorithm to estimate the robot velocity independently of wheel encoders and motor commands. It is based on the idea of fitting a motion model to a sequence of observed robot positions. In contrast to =-=[2, 3]-=-, our analysis specializes in omnidirectional robots which require the estimation of both the linear velocity as well as the angular velocity. To illustrate the benefits of velocity estimates we propo... |

1 |
K.C.: Estimation of object kinematics from point data
- Vithani, Gupta
- 2004
(Show Context)
Citation Context ...lgorithm to estimate the robot velocity independently of wheel encoders and motor commands. It is based on the idea of fitting a motion model to a sequence of observed robot positions. In contrast to =-=[2, 3]-=-, our analysis specializes in omnidirectional robots which require the estimation of both the linear velocity as well as the angular velocity. To illustrate the benefits of velocity estimates we propo... |