## Multi-robot slam with sparse extended information filters (2003)

Citations: | 51 - 4 self |

### BibTeX

@INPROCEEDINGS{Thrun03multi-robotslam,

author = {Sebastian Thrun and Yufeng Liu},

title = {Multi-robot slam with sparse extended information filters},

booktitle = {},

year = {2003},

publisher = {Springer}

}

### Years of Citing Articles

### OpenURL

### Abstract

Abstract. We present an algorithm for the multi-robot simultaneous localization and mapping (SLAM) problem. Our algorithm enables teams of robots to build joint maps, even if their relative starting locations are unknown and landmarks are ambiguous—which is presently an open problem in robotics. It achieves this capability through a sparse information filter technique, which represents maps and robot poses by Gaussian Markov random fields. The alignment of local maps into a single global maps is achieved by a tree-based algorithm for searching similar-looking local landmark configurations, paired with a hill climbing algorithm that maximizes the overall likelihood by search in the space of correspondences. We report favorable results obtained with a real-world benchmark data set. 1

### Citations

7110 |
Probabilistic reasoning in intelligent systems: networks of plausible inference
- Pearl
- 1988
(Show Context)
Citation Context ...ch removes elements in the H-matrix between a robot and a landmark without introducing new links. This step is reminiscent of the arc removal technique known from6 S. Thrun and Y. Liu Bayes networks =-=[12]-=-, with the time required for removing an arc being independent of the size of the GMRF. The sparsification is graphically illustrated in Figure 3. By removing arcs between the robot and specific landm... |

303 |
On the representation and estimation of spatial uncertainty
- Cheeseman, Smith
- 1986
(Show Context)
Citation Context ...the problem is particularly challenging for multiple robots that seek to cooperate when acquiring a map. For single-robot SLAM, the “classical” solution is based on the extended Kalman filter, or EKF =-=[14]-=-. EKFs are relatively slow when estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps [1,4,16], particle filt... |

260 | A real-time algorithm for mobile robot mapping with applications to multi-robot and 3d mapping
- Thrun, Burgard, et al.
- 2000
(Show Context)
Citation Context ...nt map. The problem is not new: A number of papers addresses the problem under the constraint that the initial pose of all robots relative to each other is known exactly [2,10,13] or in approximation =-=[5,17]-=-. All of those papers sidestep an important data association question: If two robots discover similar maps, are they actually in the same environment, or are these two different parts of the environme... |

233 | Correctness of belief propagation in Gaussian graphical models of arbitrary topology - Weiss, Freeman - 2001 |

225 | Collaborative multi-robot exploration
- Burgard, Moors, et al.
- 2000
(Show Context)
Citation Context ...e vehicles seek to build a joint map. The problem is not new: A number of papers addresses the problem under the constraint that the initial pose of all robots relative to each other is known exactly =-=[2,10,13]-=- or in approximation [5,17]. All of those papers sidestep an important data association question: If two robots discover similar maps, are they actually in the same environment, or are these two diffe... |

183 | Optimization of the simultaneous localization and map building algorithm for real time implementation, in
- Guivant, Nebot
- 2001
(Show Context)
Citation Context ...man filter, or EKF [14]. EKFs are relatively slow when estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps =-=[1,4,16]-=-, particle filters [7,9], information filters [18], and junction trees [11] into the SLAM literature. This paper addresses the topic of multi-robot SLAM, where multiple vehicles seek to build a joint ... |

168 | FastSLAM 2.0: An improved particle filtering algorithm for simultaneous localization and mapping that provably converges
- Montemerlo, Thrun, et al.
- 2003
(Show Context)
Citation Context ...Fs are relatively slow when estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps [1,4,16], particle filters =-=[7,9]-=-, information filters [18], and junction trees [11] into the SLAM literature. This paper addresses the topic of multi-robot SLAM, where multiple vehicles seek to build a joint map. The problem is not ... |

148 | An Atlas Framework for Scalable Mapping
- Bosse, Newman, et al.
- 2003
(Show Context)
Citation Context ...man filter, or EKF [14]. EKFs are relatively slow when estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps =-=[1,4,16]-=-, particle filters [7,9], information filters [18], and junction trees [11] into the SLAM literature. This paper addresses the topic of multi-robot SLAM, where multiple vehicles seek to build a joint ... |

140 | Robust mapping and localization in indoor environments using sonar data
- Tardós, Neira, et al.
(Show Context)
Citation Context ...man filter, or EKF [14]. EKFs are relatively slow when estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps =-=[1,4,16]-=-, particle filters [7,9], information filters [18], and junction trees [11] into the SLAM literature. This paper addresses the topic of multi-robot SLAM, where multiple vehicles seek to build a joint ... |

137 | Coordination for multi-robot exploration and mapping
- Simmons, Apfelbaum, et al.
- 2000
(Show Context)
Citation Context ...e vehicles seek to build a joint map. The problem is not new: A number of papers addresses the problem under the constraint that the initial pose of all robots relative to each other is known exactly =-=[2,10,13]-=- or in approximation [5,17]. All of those papers sidestep an important data association question: If two robots discover similar maps, are they actually in the same environment, or are these two diffe... |

136 | Bayesian map learning in dynamic environments
- Murphy
- 1999
(Show Context)
Citation Context ...Fs are relatively slow when estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps [1,4,16], particle filters =-=[7,9]-=-, information filters [18], and junction trees [11] into the SLAM literature. This paper addresses the topic of multi-robot SLAM, where multiple vehicles seek to build a joint map. The problem is not ... |

122 | Thin Junction Tree Filters for Simultaneous Localization and
- Paskin
- 2003
(Show Context)
Citation Context ...eading direction, which is the very first element in the state vector. The heading simply changes by the angle of the rotation between both maps, denoted α in (12). To see the correctness of (10) and =-=(11)-=-, we recall that the parameters 〈H j t , b j t〉 define a Gaussian over the j-th robot pose ans map x j t = ( x j t Y ) T . This gives us the following derivation: p(x j t | Z j , U j ) (11)8 S. Thrun... |

78 | A solution to the simultaneous localisation and map building (slam) problem
- Dissanayake, Newman, et al.
- 2001
(Show Context)
Citation Context ...ts obtained with a real-world benchmark data set. 1 Introduction Simultaneous localization and mapping, or SLAM, addresses the problem of acquiring an environment map with one or more moving vehicles =-=[3]-=-. In statistical terms, SLAM is a high-dimensional estimation problem characterized by two major sources of uncertainty, pertaining to the noise in sensing and in motion. Whereas SLAM has mostly been ... |

51 |
An experimental system for incremental environment modeling by an autonomous mobile robot
- Moutarlier, Chatila
- 1989
(Show Context)
Citation Context ...obots from time 0 to time t; U is the set of all controls. 3 Sparse Extended Information Filters for Multi-Robot Systems The classical solution to the SLAM problem is the extended Kalman filter (EKF) =-=[8,14]-=-. The EKF approximates the posterior p(Xt, Y | Z, U) by a multivariate Gaussian, with mean vector µt and covariance Σt. Updating this Gaussian is achieved by linearizing g and h at µt, and applying th... |

47 | Simultaneous mapping and localization with sparse extended information filters
- Thrun
(Show Context)
Citation Context ... estimating high-dimensional maps. Recent research has led to a flurry of more capable algorithms, introducing concepts such as hierarchical maps [1,4,16], particle filters [7,9], information filters =-=[18]-=-, and junction trees [11] into the SLAM literature. This paper addresses the topic of multi-robot SLAM, where multiple vehicles seek to build a joint map. The problem is not new: A number of papers ad... |

10 |
Multiple platform localisation and map building
- Nettleton, Durrant-Whyte, et al.
- 2000
(Show Context)
Citation Context ...e vehicles seek to build a joint map. The problem is not new: A number of papers addresses the problem under the constraint that the initial pose of all robots relative to each other is known exactly =-=[2,10,13]-=- or in approximation [5,17]. All of those papers sidestep an important data association question: If two robots discover similar maps, are they actually in the same environment, or are these two diffe... |

8 |
Mobile robot sense net
- Konolige, Gutmann, et al.
- 1999
(Show Context)
Citation Context ...nt map. The problem is not new: A number of papers addresses the problem under the constraint that the initial pose of all robots relative to each other is known exactly [2,10,13] or in approximation =-=[5,17]-=-. All of those papers sidestep an important data association question: If two robots discover similar maps, are they actually in the same environment, or are these two different parts of the environme... |

7 |
A hierarchical bayesian approach to mobile robot map structure estimation
- Stewart, Ko, et al.
- 2003
(Show Context)
Citation Context ...rt at the same location. If their initial location is unknown—which is the case addressed in this paper—it creates a challenging data association problem. This problem was addressed is a recent paper =-=[15]-=-: the idea here is that robots continuously attempt to localize themselves in each other’s maps using particle filters. While this is a highly promising approach, it is computationally somewhat expens... |