US 12,476,880 B2
Method and apparatus for predictive maintenance
Armen Aghasaryan, Massy (FR); Kerim Palamutcuogullari, Reading (GB); Rajesh Banerjee, Dallas, TX (US); and Dimitre Davidov Kostadinov, Massy (FR)
Assigned to Nokia Solutions and Networks Oy, Espoo (FI)
Filed by Nokia Solutions and Networks Oy, Espoo (FI)
Filed on Jan. 10, 2024, as Appl. No. 18/408,709.
Claims priority of application No. 20235061 (FI), filed on Jan. 20, 2023.
Prior Publication US 2024/0250882 A1, Jul. 25, 2024
Int. Cl. G06F 15/173 (2006.01); H04L 41/147 (2022.01); H04L 41/16 (2022.01)
CPC H04L 41/147 (2013.01) [H04L 41/16 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus for predictive maintenance of modules in a telecommunications network, the apparatus comprising:
at least one processor; and
at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the apparatus at least to perform:
collecting a primary data bundle, wherein the primary data bundle comprises a first set of measurement time-series generated by the telecommunications network during a first time-window, wherein the measurement time-series of the first set is generated by a plurality of network elements of the telecommunication network and relate to a same physical module comprising one or more hardware components in the network elements that are subject to maintenance operations,
processing maintenance data relating to maintenance of the physical module or to maintenance of a network element comprising the physical module, to determine whether the primary data bundle is associated with a fault-related maintenance intervention on the physical module carried out after the first time-window, wherein the maintenance data comprises records of on-field replacement of the physical module in the telecommunications network,
in response to determining that the primary data bundle is associated with a maintenance intervention on the physical module, creating secondary data bundles comprising a second set of measurement time-series generated by the telecommunications network during a second time-window started before the first time-window, wherein the measurement time-series of the second set relate to the physical module of the telecommunications network, wherein the secondary data bundles cover distinct consecutive time periods of a same duration, and
training a predictive model using the primary data bundle and the secondary data bundles, wherein the predictive model is configured to compute an anomaly-related prediction relating to the telecommunications network.