## Elastic correction of dead-reckoning errors in map building (1998)

Venue: | In Intl. Conf. on Intelligent Robots and Systems |

Citations: | 36 - 1 self |

### BibTeX

@INPROCEEDINGS{Golfarelli98elasticcorrection,

author = {Matteo Golfarelli and Dario Maio and Stefano Rizzi},

title = {Elastic correction of dead-reckoning errors in map building},

booktitle = {In Intl. Conf. on Intelligent Robots and Systems},

year = {1998},

pages = {905--911}

}

### Years of Citing Articles

### OpenURL

### Abstract

Abstract—Map building is an important issue for all the applications in mobile robotics in which the environment is unknown and, in general, in order to have a robot exhibit a fully autonomous behavior. A major problem in map building is due to the imprecision of sensor measures. In this paper, we propose a technique, called elastic correction, for correcting the dead-reckoning errors made during the exploration of an environment by a robot capable of identifying landmarks. Knowledge being acquired is modeled by a relational graph whose vertices and arcs represent, respectively, landmarks and routes. Elastic correction is based on an analogy between the graph modeling the environment and a mechanical structure: the map is regarded as a truss where each route is an elastic bar and each landmark a node. Errors are corrected as a result of the deformations induced from the forces arising within the structure as inconsistent measures are taken. The elasticity parameters characterizing the structure are used to model the uncertainty on odometry. The paper presents results from simulations showing the effectiveness of the method for reducing the overall metric error and proving its robustness with reference to topological errors and to unpredictable sensor errors. Index Terms—Error correction, mobile robotics, odometry. I.

### Citations

1081 |
Practical Methods of Optimization
- Fletcher
- 1981
(Show Context)
Citation Context ...ing and can be formulated as a constrained optimization problem where the function to be minimized is while the constraint is expressed by (15). This problem can be solved using the Lagrangian method =-=[20]-=- determining the linear system (16) (17) that admits an analytic solution. Fig. 7 shows the error made by the robot while covering an open polygon, how the inconsistency is solved and the correspondin... |

427 | Globally consistent range scan alignment for environment mapping. Autonomous Robots, 1997. BIBLIOGRAPHIE tel-00545774, version 1 - 12 Dec 2010 [LM97b] Feng Lu and Evangelos Milios. Robot pose estimation in unknown environments by matching 2d range scans
- Lu, Milios
- 1997
(Show Context)
Citation Context ...oth the uncertainty on odometry and the measurements collected so far; this enables the exploitation of previously collected knowledge. Two other methods sharing this feature with EC are described in =-=[14]-=- and [15], and will be analyzed in Section III-E. The paper is organized as follows. Section II reports our assumptions concerning the navigation model. Section III outlines the necessary background o... |

403 | A probabilistic approach to concurrent mapping and localization for mobile robots
- Thrun, Fox, et al.
- 1998
(Show Context)
Citation Context ...akes only metric errors into account. Unfortunately, the map building process may be heavily affected by topological errors. Due to their relevance several works discuss this issue (for instance, see =-=[15]-=-). Although extending EC in order to correct topological errors is beyond the scope of this paper, we conducted some tests aimed at showing how robust EC is to them. Within our graph-based framework, ... |

289 |
Sonar-based real-world mapping and navigation
- Elfes
- 1987
(Show Context)
Citation Context .... Several techniques for environment representation have been devised in the literature, including topological maps [9], symbolic graphs labelled with metric information [2] and analogic descriptions =-=[6]-=-. Though in some applications a detailed map of the environment is given to the robot a priori, map building is still an important issue for all the applications in which the environment is unknown an... |

150 |
T.Weymouth, “Topological mapping for mobile robots using a combination of sonar and vision sensing
- Kortenkamp
- 1994
(Show Context)
Citation Context ...of the environment to carry out successfully the navigational tasks assigned to them. Several techniques for environment representation have been devised in the literature, including topological maps =-=[9]-=-, symbolic graphs labelled with metric information [2] and analogic descriptions [6]. Though in some applications a detailed map of the environment is given to the robot a priori, map building is stil... |

133 | Learning topological maps with weak local odometric information
- Shatkay, Kaelbling
- 1997
(Show Context)
Citation Context ...methods proposed in the literature can hardly be quantitatively compared with elastical correction, since they are either based on assumptions and sensory equipments radically different from ours [7] =-=[13]-=- or were devised within a different framework [14]. Figure 7. First-sight correction. The measured map is grey, the corrected one is black; crosses represent the true positions of vertices. Figure 7 a... |

131 |
A distributed model for mobile robot environment-learning and navigation
- Mataric
- 1990
(Show Context)
Citation Context ... The elasticity parameters characterizing the structure are used to model the uncertainty on odometry. The main advantage of our approach as compared to those proposed in the literature (e.g. [2] [7] =-=[12]-=-) is that it takes into account all the information previously collected in order to evaluate the correctness of the new measurements being acquired. 2 Uncertainty in Self-Positioning Let the pose of ... |

118 | Measurement and correction of systematic odometry errors in mobile robots
- Borenstein, Feng
- 1996
(Show Context)
Citation Context ...recision of measurements, which produces metric errors. In particular, odometers typically produce both systematic and non-systematic errors. The former depend entirely on the mobile platform adopted =-=[4]-=-; the latter are due to undesired interactions between the robot and the environment, such as uneven ground. Systematic errors can be predicted; some are deterministic, some can be modelled by a proba... |

117 | Mobile robot localization using landmarks
- Betke, Gurvits
- 1993
(Show Context)
Citation Context ...[6] or on the possibility of recognizing landmarks whose position in the environment is known; landmarks may be identified by processing data obtained by sensors such as sonars, lasers, cameras, etc. =-=[7]-=-, [8]. Problems influencing the effectiveness of these techniques in the real world are caused by the imprecision of measurements, which produces metric errors. In particular, odometers typically prod... |

81 | Unsupervised Learning of Probabilistic Models for Robot Navigation
- Koenig, Simmons
- 1996
(Show Context)
Citation Context ...red route and consist in a change of the link confidence; while calculating the new confidence value, the algorithm does not take any kind of information about the surrounding routes into account. In =-=[24]-=-, the authors propose a technique to learn metric information about the environment, which requires a priori topological knowledge of the environment. IV. EXPERIMENTAL RESULTS We have extensively test... |

75 | Spatial learning for navigation in dynamic environments
- Yamauchi, Beer
- 1996
(Show Context)
Citation Context ...quantitatively compared with EC, since they are either based on assumptions and sensory equipment radically different from ours [12] or were devised within a different framework [22]. For example, in =-=[23]-=-, the authors present an algorithm based on an adaptive place network (APN) containing both metric and topological information about the structure of the environment. The APN uses no landmarks to repr... |

61 |
Error correction in mobile robot map learning
- Engelson, McDermott
- 1992
(Show Context)
Citation Context ...ken. The elasticity parameters characterizing the structure are used to model the uncertainty on odometry. The main advantage of our approach as compared to those proposed in the literature (e.g. [2] =-=[7]-=- [12]) is that it takes into account all the information previously collected in order to evaluate the correctness of the new measurements being acquired. 2 Uncertainty in Self-Positioning Let the pos... |

31 | Coping with uncertainty in map learning
- Basye, Vitter
- 1997
(Show Context)
Citation Context ...ational tasks assigned to them. Several techniques for environment representation have been devised in the literature, including topological maps [9], symbolic graphs labelled with metric information =-=[2]-=- and analogic descriptions [6]. Though in some applications a detailed map of the environment is given to the robot a priori, map building is still an important issue for all the applications in which... |

27 |
Mobile robot self-location using model-image feature correspondence
- Talluri, Aggarwal
- 1996
(Show Context)
Citation Context ... topological maps [1], symbolic graphs labeled with metric information [2], and analogic descriptions [3]. Though in some applications a detailed map of the environment is given to the robot a priori =-=[4]-=-, map building is still an important issue for all the applications in which the environment is unknown (e.g., undersea and space exploration) and, in general, in order to have a robot exhibit a fully... |

17 | Exploration strategies for mobile robots, in
- Taylor, Kdegman
- 1993
(Show Context)
Citation Context .... Several techniques for environment representation have been devised in the literature, including topological maps [1], symbolic graphs labeled with metric information [2], and analogic descriptions =-=[3]-=-. Though in some applications a detailed map of the environment is given to the robot a priori [4], map building is still an important issue for all the applications in which the environment is unknow... |

13 | The CLAPPER: A Dual-drive Mobile Robot With Internal Correction of Dead-reckoning Errors
- Borenstein
- 1994
(Show Context)
Citation Context ...r; we may assume that they are distributed normally with null mean. Non-systematic errors cannot be predicted, unless some assumptions on the environment can be made or particular sensors are mounted =-=[3]-=-. Our approach to map building can be applied to correct the dead-reckoning errors made by a robot navigating within an environment where landmarks are present; we do not consider topological errors s... |

13 |
Probabilistic map learning: Necessity and difficulties
- Hébert, Betgé-Brezetz, et al.
- 1995
(Show Context)
Citation Context ...is, the less effective error correction is. In the tests shown in Section IV we used . E. Related Literature Among the error correction techniques in the literature, those proposed in [15], [14], and =-=[21]-=- (which we will call, respectively, TH, ML, and HE) share with EC the capability of taking all previously collected measurements into account and, at the same time, that of using locally acquired know... |

12 |
Fusion of Deadreckoning Positions with a Workspace Model for a Mobile Robot by Bayesian Inference
- Tanoushi, Tsubouchi, et al.
(Show Context)
Citation Context ...) are still calculated as above, j (k+1) may be measured directly; thus, each new positional estimate is not affected by the errors made in measuring the robot's orientation at the previous steps. In =-=[15]-=- an estimate of the positional uncertainty for a robot moving along a path and calculating its current position by odometers is derived. Each position calculated is associated with a density function ... |

11 |
Non-Uniform Dead-Reckoning Position Estimate Updates
- Krantz
- 1996
(Show Context)
Citation Context ... absolute. Relative positioning is based on odometers which estimate the current position by determining the offset from the initial position, computed by counting the revolutions of the robot wheels =-=[5]-=-. Absolute positioning is based either on an absolute sensor such as a GPS [6] or on the possibility of recognizing landmarks whose position in the environment is known; landmarks may be identified by... |

9 |
Dynamic Clustering Of Maps In Autonomous Agents
- Maio, Maltoni, et al.
- 1996
(Show Context)
Citation Context ...nizing a landmark it has already met. The environment is modelled by a relational graph whose vertices and arcs represent, respectively, the landmarks sensed and the inter-landmark routes experienced =-=[10]-=-. While exploring the environment, the robot calculates the relative position of each landmark compared to the one met immediately before, by applying dead-reckoning; when it meets a landmark it has a... |

7 |
Mobile robot localization in indoor environment
- Dulimatra, Jain
- 1997
(Show Context)
Citation Context ...r on the possibility of recognizing landmarks whose position in the environment is known; landmarks may be identified by processing data obtained by sensors such as sonars, lasers, cameras, etc. [7], =-=[8]-=-. Problems influencing the effectiveness of these techniques in the real world are caused by the imprecision of measurements, which produces metric errors. In particular, odometers typically produce b... |

5 |
C.: Introduction to MatrixMethods of StructuralAnalbsis, McGraw-Hill, _L
- MARTIN
- 1966
(Show Context)
Citation Context ...le structure is assembled by superimposing the stiffness matrices of the single elements belonging to the structure. The displacements of the nodes can then be solved by applying the stiffness method =-=[11]-=-. 3.2 Correction of first-sight errors Suppose the robot is exploring an unknown area starting from a known landmark v 0 . It meets the unknown landmarks v 1 ,...v m-1 by covering a sequence of unknow... |

4 |
A Method for an Autonomous Mobile Robot to Recognize its Position in the Global Coordinate System when Building a
- Sugiyama
(Show Context)
Citation Context ...uantitatively compared with elastical correction, since they are either based on assumptions and sensory equipments radically different from ours [7] [13] or were devised within a different framework =-=[14]-=-. Figure 7. First-sight correction. The measured map is grey, the corrected one is black; crosses represent the true positions of vertices. Figure 7 and Table I show the result of first-sight correcti... |

3 |
A hierarchical approach to sonar-based landmark detection in mobile robots
- Golfarelli, Maio, et al.
- 1997
(Show Context)
Citation Context ...bsolute positioning is based either on an absolute sensor such as a GPS [5] or on the possibility of recognizing landmarks by processing data obtained by sensors such as sonars, lasers, cameras, etc. =-=[8]-=-. Problems influencing the effectiveness of these techniques in the real world are caused by the imprecision of measurements, which produces metric errors. In particular, odometers typically produce b... |

3 |
et al, "Control and Localisation of a Post Distributing Mobile Robot
- Adams
(Show Context)
Citation Context ...in which the robot may stand with nonnegligible probability is an ellipse whose shape and dimensions depend on the length and complexity of the path, and is represented by the covariance matrix C (k) =-=[1]-=-. Similarly, an estimate of the probability density function of the position for a robot mounting a compass can be derived. 3 Elastic Correction The map built by the robot is structured as a nondirect... |

2 |
A Kalman filter for GPS navigation of land vehicles
- Cooper, Durrant-Whyte
- 1994
(Show Context)
Citation Context ...ive positioning is based on odometers which estimate the current position by determining the offset from the initial position. Absolute positioning is based either on an absolute sensor such as a GPS =-=[5]-=- or on the possibility of recognizing landmarks by processing data obtained by sensors such as sonars, lasers, cameras, etc. [8]. Problems influencing the effectiveness of these techniques in the real... |

1 |
et al., "Control and localisation of a post distributing mobile robot
- Adams
- 1994
(Show Context)
Citation Context ...in which the robot may stand with nonnegligible probability is an ellipse whose shape and dimensions depend on the length and complexity of the path, and is represented by the covariance matrix C (k) =-=[1]-=-. Similarly, an estimate of the probability density function of the position for a robot mounting a compass can be derived. 3 Elastic Correction The map built by the robot is structured as a nondirect... |

1 |
et al., “Control and localization of a post distributing mobile robot
- Adams
- 1994
(Show Context)
Citation Context ...t is positioned in the surrounding area. The area in which the robot may stand with non-negligible probability is an ellipse whose shape and dimensions depend on the length and complexity of the path =-=[17]-=-. In the Appendix, we derive an estimate of the probability density function of the position for a robot mounting a compass. While from a theoretical point of view every density function could be used... |

1 |
et al., “Map learning and high-speed navigation
- Thrun
- 1998
(Show Context)
Citation Context ... an estimate of the probability density function of the position for a robot mounting a compass. While from a theoretical point of view every density function could be used, previous experience [14], =-=[18]-=-, [16] shows that the errors can be effectively modeled using a Gaussian distribution with null mean [3]. In this case, the uncertainty on the robot position is distributed around the position calcula... |