Robust, low-bandwidth, multi-vehicle mapping
This paper addresses the problem of decentralised simultaneous localisation and map building for a team of agents where the communication bandwidth is limited. We present an extension to current approaches that enables multiple vehicles to acquire a joint map, but which can cope with communication b...
| Main Authors: | , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
2005
|
| _version_ | 1797063303935033344 |
|---|---|
| author | Reece, S Roberts, S IEEE |
| author_facet | Reece, S Roberts, S IEEE |
| author_sort | Reece, S |
| collection | OXFORD |
| description | This paper addresses the problem of decentralised simultaneous localisation and map building for a team of agents where the communication bandwidth is limited. We present an extension to current approaches that enables multiple vehicles to acquire a joint map, but which can cope with communication bandwidth limitations. Nettleton's approach uses a hybrid information filter/Covariance Intersection algorithm on each communication link to manage the inter-vehicle communication and ensure that information vehicles share does not get 'double counted'. The Covariance Intersection algorithm is a highly conservative method for managing double counting and its use can produce highly uncertain maps. We introduce a novel and more efficient tool, called Bounded Covariance Inflation, for managing the double counting (or rumour propagation) problem. We show that the parameters required by the new approach can be determined locally by each vehicle and therefore the decentralised nature of the network is not compromised. We provide experimental results that illustrate the effectiveness of our approach in comparison with the original approach of Nettleton et al. © 2005 IEEE. |
| first_indexed | 2024-03-06T20:57:55Z |
| format | Journal article |
| id | oxford-uuid:39e2a7ed-e897-4a38-9125-7e78109eb609 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T20:57:55Z |
| publishDate | 2005 |
| record_format | dspace |
| spelling | oxford-uuid:39e2a7ed-e897-4a38-9125-7e78109eb6092022-03-26T13:58:11ZRobust, low-bandwidth, multi-vehicle mappingJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:39e2a7ed-e897-4a38-9125-7e78109eb609EnglishSymplectic Elements at Oxford2005Reece, SRoberts, SIEEEThis paper addresses the problem of decentralised simultaneous localisation and map building for a team of agents where the communication bandwidth is limited. We present an extension to current approaches that enables multiple vehicles to acquire a joint map, but which can cope with communication bandwidth limitations. Nettleton's approach uses a hybrid information filter/Covariance Intersection algorithm on each communication link to manage the inter-vehicle communication and ensure that information vehicles share does not get 'double counted'. The Covariance Intersection algorithm is a highly conservative method for managing double counting and its use can produce highly uncertain maps. We introduce a novel and more efficient tool, called Bounded Covariance Inflation, for managing the double counting (or rumour propagation) problem. We show that the parameters required by the new approach can be determined locally by each vehicle and therefore the decentralised nature of the network is not compromised. We provide experimental results that illustrate the effectiveness of our approach in comparison with the original approach of Nettleton et al. © 2005 IEEE. |
| spellingShingle | Reece, S Roberts, S IEEE Robust, low-bandwidth, multi-vehicle mapping |
| title | Robust, low-bandwidth, multi-vehicle mapping |
| title_full | Robust, low-bandwidth, multi-vehicle mapping |
| title_fullStr | Robust, low-bandwidth, multi-vehicle mapping |
| title_full_unstemmed | Robust, low-bandwidth, multi-vehicle mapping |
| title_short | Robust, low-bandwidth, multi-vehicle mapping |
| title_sort | robust low bandwidth multi vehicle mapping |
| work_keys_str_mv | AT reeces robustlowbandwidthmultivehiclemapping AT robertss robustlowbandwidthmultivehiclemapping AT ieee robustlowbandwidthmultivehiclemapping |